{
  "schemaVersion": 2,
  "phase": "scored",
  "synthetic": true,
  "status": "complete",
  "matrixComplete": true,
  "models": [
    "openai/gpt-5.6-luna",
    "openai/gpt-5.6-sol",
    "anthropic/claude-opus-5",
    "anthropic/claude-sonnet-5",
    "deepseek/deepseek-v4-flash-0731",
    "deepseek/deepseek-v4-pro-0813",
    "alibaba/qwen3.8-flash",
    "alibaba/qwen3.8-max-0902",
    "google/gemini-3.8-flash",
    "openai/gpt-6-astra",
    "zai/glm-5.3-flash"
  ],
  "excludedModels": [
    {
      "model": "meta/muse-spark-1.3",
      "reason": "Excluded from the main study at the user’s request because one timeout charge is unresolved. Original runs retained."
    }
  ],
  "sourceObservedRuns": 1080,
  "expectedRuns": 990,
  "observedRuns": 990,
  "groups": [
    {
      "model": "openai/gpt-5.6-luna",
      "workflow": "overall",
      "attempts": 90,
      "successes": 88,
      "successRate": 0.9777777777777777,
      "billingComplete": true,
      "totalCost": 0.09424678999999997,
      "costPerAttempt": 0.0010471865555555552,
      "costPerSuccess": 0.0010709862499999996,
      "successfulSpend": 0.09204421999999997,
      "failedSpend": 0.00220257,
      "latencyMs": 8933.044444444444,
      "inputTokens": 117947,
      "outputTokens": 27053,
      "reasoningTokens": 8288,
      "cacheReadTokens": 182047,
      "cacheWriteTokens": 98429,
      "toolCalls": 378,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          0.9444444444444444,
          1
        ],
        "costPerSuccess": [
          0.0010142252808988763,
          0.0011386378160919536
        ]
      }
    },
    {
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "attempts": 30,
      "successes": 29,
      "successRate": 0.9666666666666667,
      "billingComplete": true,
      "totalCost": 0.02949795,
      "costPerAttempt": 0.0009832649999999999,
      "costPerSuccess": 0.0010171706896551723,
      "successfulSpend": 0.028441250000000005,
      "failedSpend": 0.0010566999999999998,
      "latencyMs": 8774.933333333332,
      "inputTokens": 43979,
      "outputTokens": 8432,
      "reasoningTokens": 2259,
      "cacheReadTokens": 50035,
      "cacheWriteTokens": 27489,
      "toolCalls": 96,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          0.9,
          1
        ],
        "costPerSuccess": [
          0.000896916666666667,
          0.0011961570370370373
        ]
      }
    },
    {
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "attempts": 30,
      "successes": 30,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 0.031035749999999994,
      "costPerAttempt": 0.001034525,
      "costPerSuccess": 0.001034525,
      "successfulSpend": 0.031035749999999994,
      "failedSpend": 0,
      "latencyMs": 8773.833333333334,
      "inputTokens": 51036,
      "outputTokens": 8533,
      "reasoningTokens": 2407,
      "cacheReadTokens": 38740,
      "cacheWriteTokens": 27703,
      "toolCalls": 126,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.000980605333333333,
          0.0010891063333333332
        ]
      }
    },
    {
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "attempts": 30,
      "successes": 29,
      "successRate": 0.9666666666666667,
      "billingComplete": true,
      "totalCost": 0.033713089999999994,
      "costPerAttempt": 0.0011237696666666665,
      "costPerSuccess": 0.001162520344827586,
      "successfulSpend": 0.032567219999999994,
      "failedSpend": 0.00114587,
      "latencyMs": 9250.366666666667,
      "inputTokens": 22932,
      "outputTokens": 10088,
      "reasoningTokens": 3622,
      "cacheReadTokens": 93272,
      "cacheWriteTokens": 43237,
      "toolCalls": 156,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          0.9,
          1
        ],
        "costPerSuccess": [
          0.0010721244827586209,
          0.0012774203703703703
        ]
      }
    },
    {
      "model": "openai/gpt-5.6-sol",
      "workflow": "overall",
      "attempts": 90,
      "successes": 90,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 0.8361001,
      "costPerAttempt": 0.00929000111111111,
      "costPerSuccess": 0.00929000111111111,
      "successfulSpend": 0.8361001,
      "failedSpend": 0,
      "latencyMs": 10298.9,
      "inputTokens": 122703,
      "outputTokens": 26317,
      "reasoningTokens": 3987,
      "cacheReadTokens": 168313,
      "cacheWriteTokens": 101579,
      "toolCalls": 387,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.008942644444444441,
          0.009688347777777771
        ]
      }
    },
    {
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "attempts": 30,
      "successes": 30,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 0.26205490000000004,
      "costPerAttempt": 0.008735163333333334,
      "costPerSuccess": 0.008735163333333334,
      "successfulSpend": 0.26205490000000004,
      "failedSpend": 0,
      "latencyMs": 9198.733333333334,
      "inputTokens": 44927,
      "outputTokens": 8072,
      "reasoningTokens": 1024,
      "cacheReadTokens": 43242,
      "cacheWriteTokens": 29037,
      "toolCalls": 96,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.008134690000000002,
          0.009346460000000003
        ]
      }
    },
    {
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "attempts": 30,
      "successes": 30,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 0.29374229999999996,
      "costPerAttempt": 0.009791409999999999,
      "costPerSuccess": 0.009791409999999999,
      "successfulSpend": 0.29374229999999996,
      "failedSpend": 0,
      "latencyMs": 10028.166666666666,
      "inputTokens": 54857,
      "outputTokens": 8787,
      "reasoningTokens": 1297,
      "cacheReadTokens": 42354,
      "cacheWriteTokens": 29887,
      "toolCalls": 135,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.009397183333333333,
          0.010201893333333333
        ]
      }
    },
    {
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "attempts": 30,
      "successes": 30,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 0.28030289999999997,
      "costPerAttempt": 0.009343429999999998,
      "costPerSuccess": 0.009343429999999998,
      "successfulSpend": 0.28030289999999997,
      "failedSpend": 0,
      "latencyMs": 11669.8,
      "inputTokens": 22919,
      "outputTokens": 9458,
      "reasoningTokens": 1666,
      "cacheReadTokens": 82717,
      "cacheWriteTokens": 42655,
      "toolCalls": 156,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.008571206666666664,
          0.010269173333333331
        ]
      }
    },
    {
      "model": "anthropic/claude-opus-5",
      "workflow": "overall",
      "attempts": 90,
      "successes": 89,
      "successRate": 0.9888888888888889,
      "billingComplete": true,
      "totalCost": 3.0610752500000005,
      "costPerAttempt": 0.03401194722222223,
      "costPerSuccess": 0.034394103932584276,
      "successfulSpend": 2.9962935000000006,
      "failedSpend": 0.06478175,
      "latencyMs": 12443.255555555555,
      "inputTokens": 604,
      "outputTokens": 62366,
      "reasoningTokens": 5308,
      "cacheReadTokens": 547873,
      "cacheWriteTokens": 174763,
      "toolCalls": 435,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          0.9666666666666667,
          1
        ],
        "costPerSuccess": [
          0.031949230555555534,
          0.037141790229885056
        ]
      }
    },
    {
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "attempts": 30,
      "successes": 30,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 0.8715719999999999,
      "costPerAttempt": 0.029052399999999996,
      "costPerSuccess": 0.029052399999999996,
      "successfulSpend": 0.8715719999999999,
      "failedSpend": 0,
      "latencyMs": 11679.1,
      "inputTokens": 202,
      "outputTokens": 17500,
      "reasoningTokens": 1146,
      "cacheReadTokens": 177224,
      "cacheWriteTokens": 50528,
      "toolCalls": 125,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.02490031666666666,
          0.03281325833333334
        ]
      }
    },
    {
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "attempts": 30,
      "successes": 30,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 0.9212950000000001,
      "costPerAttempt": 0.030709833333333336,
      "costPerSuccess": 0.030709833333333336,
      "successfulSpend": 0.9212950000000001,
      "failedSpend": 0,
      "latencyMs": 11056.3,
      "inputTokens": 206,
      "outputTokens": 18770,
      "reasoningTokens": 1530,
      "cacheReadTokens": 177355,
      "cacheWriteTokens": 51854,
      "toolCalls": 147,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.026922125000000005,
          0.03458180833333333
        ]
      }
    },
    {
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "attempts": 30,
      "successes": 29,
      "successRate": 0.9666666666666667,
      "billingComplete": true,
      "totalCost": 1.2682082499999998,
      "costPerAttempt": 0.04227360833333333,
      "costPerSuccess": 0.043731318965517234,
      "successfulSpend": 1.2034264999999997,
      "failedSpend": 0.06478175,
      "latencyMs": 14594.366666666667,
      "inputTokens": 196,
      "outputTokens": 26096,
      "reasoningTokens": 2632,
      "cacheReadTokens": 193294,
      "cacheWriteTokens": 72381,
      "toolCalls": 163,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          0.9,
          1
        ],
        "costPerSuccess": [
          0.03882839166666666,
          0.051619268518518516
        ]
      }
    },
    {
      "model": "anthropic/claude-sonnet-5",
      "workflow": "overall",
      "attempts": 90,
      "successes": 87,
      "successRate": 0.9666666666666667,
      "billingComplete": true,
      "totalCost": 1.4137248000000002,
      "costPerAttempt": 0.015708053333333336,
      "costPerSuccess": 0.016249710344827587,
      "successfulSpend": 1.3514370000000002,
      "failedSpend": 0.062287800000000004,
      "latencyMs": 12626.266666666666,
      "inputTokens": 556,
      "outputTokens": 58041,
      "reasoningTokens": 11637,
      "cacheReadTokens": 423114,
      "cacheWriteTokens": 252484,
      "toolCalls": 384,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          0.9111111111111111,
          1
        ],
        "costPerSuccess": [
          0.014745524444444442,
          0.018041012790697676
        ]
      }
    },
    {
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "attempts": 30,
      "successes": 30,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 0.4073892,
      "costPerAttempt": 0.01357964,
      "costPerSuccess": 0.01357964,
      "successfulSpend": 0.4073892,
      "failedSpend": 0,
      "latencyMs": 11112.6,
      "inputTokens": 182,
      "outputTokens": 17032,
      "reasoningTokens": 2402,
      "cacheReadTokens": 131001,
      "cacheWriteTokens": 74594,
      "toolCalls": 97,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.012094240000000001,
          0.014973939999999998
        ]
      }
    },
    {
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "attempts": 30,
      "successes": 29,
      "successRate": 0.9666666666666667,
      "billingComplete": true,
      "totalCost": 0.4315710000000001,
      "costPerAttempt": 0.014385700000000003,
      "costPerSuccess": 0.014881758620689658,
      "successfulSpend": 0.4075153000000001,
      "failedSpend": 0.0240557,
      "latencyMs": 12101.366666666667,
      "inputTokens": 184,
      "outputTokens": 17787,
      "reasoningTokens": 3447,
      "cacheReadTokens": 130115,
      "cacheWriteTokens": 77136,
      "toolCalls": 130,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          0.9,
          1
        ],
        "costPerSuccess": [
          0.012296733333333334,
          0.018924196296296294
        ]
      }
    },
    {
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "attempts": 30,
      "successes": 28,
      "successRate": 0.9333333333333333,
      "billingComplete": true,
      "totalCost": 0.5747646,
      "costPerAttempt": 0.01915882,
      "costPerSuccess": 0.02052730714285714,
      "successfulSpend": 0.5365324999999999,
      "failedSpend": 0.038232100000000005,
      "latencyMs": 14664.833333333334,
      "inputTokens": 190,
      "outputTokens": 23222,
      "reasoningTokens": 5788,
      "cacheReadTokens": 161998,
      "cacheWriteTokens": 100754,
      "toolCalls": 157,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          0.8,
          1
        ],
        "costPerSuccess": [
          0.01730070333333333,
          0.025205696153846158
        ]
      }
    },
    {
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "overall",
      "attempts": 90,
      "successes": 89,
      "successRate": 0.9888888888888889,
      "billingComplete": true,
      "totalCost": 0.02464238799999999,
      "costPerAttempt": 0.000273804311111111,
      "costPerSuccess": 0.0002768807640449437,
      "successfulSpend": 0.02413856499999999,
      "failedSpend": 0.000503823,
      "latencyMs": 6867.455555555555,
      "inputTokens": 101601,
      "outputTokens": 69576,
      "reasoningTokens": 0,
      "cacheReadTokens": 448256,
      "cacheWriteTokens": 0,
      "toolCalls": 473,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          0.9666666666666667,
          1
        ],
        "costPerSuccess": [
          0.0002582635,
          0.0002964073820224719
        ]
      }
    },
    {
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "attempts": 30,
      "successes": 30,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 0.007455158999999999,
      "costPerAttempt": 0.0002485053,
      "costPerSuccess": 0.0002485053,
      "successfulSpend": 0.007455158999999999,
      "failedSpend": 0,
      "latencyMs": 6409.2,
      "inputTokens": 31038,
      "outputTokens": 20519,
      "reasoningTokens": 0,
      "cacheReadTokens": 139776,
      "cacheWriteTokens": 0,
      "toolCalls": 133,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.00022739103333333332,
          0.00026634146666666664
        ]
      }
    },
    {
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "attempts": 30,
      "successes": 30,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 0.007286694000000001,
      "costPerAttempt": 0.0002428898,
      "costPerSuccess": 0.0002428898,
      "successfulSpend": 0.007286694000000001,
      "failedSpend": 0,
      "latencyMs": 6785.8,
      "inputTokens": 28621,
      "outputTokens": 20314,
      "reasoningTokens": 0,
      "cacheReadTokens": 143104,
      "cacheWriteTokens": 0,
      "toolCalls": 159,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.00021280333333333334,
          0.0002729345000000001
        ]
      }
    },
    {
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "attempts": 30,
      "successes": 29,
      "successRate": 0.9666666666666667,
      "billingComplete": true,
      "totalCost": 0.009900535,
      "costPerAttempt": 0.0003300178333333333,
      "costPerSuccess": 0.00034139775862068966,
      "successfulSpend": 0.009396712000000002,
      "failedSpend": 0.000503823,
      "latencyMs": 7407.366666666667,
      "inputTokens": 41942,
      "outputTokens": 28743,
      "reasoningTokens": 0,
      "cacheReadTokens": 165376,
      "cacheWriteTokens": 0,
      "toolCalls": 181,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          0.9,
          1
        ],
        "costPerSuccess": [
          0.0002958168666666666,
          0.00039331124137931045
        ]
      }
    },
    {
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "overall",
      "attempts": 90,
      "successes": 89,
      "successRate": 0.9888888888888889,
      "billingComplete": true,
      "totalCost": 0.5746738359999999,
      "costPerAttempt": 0.006385264844444443,
      "costPerSuccess": 0.0064570093932584255,
      "successfulSpend": 0.5705111279999999,
      "failedSpend": 0.004162708,
      "latencyMs": 12778.844444444445,
      "inputTokens": 247552,
      "outputTokens": 42702,
      "reasoningTokens": 17068,
      "cacheReadTokens": 254909,
      "cacheWriteTokens": 0,
      "toolCalls": 442,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          0.9666666666666667,
          1
        ],
        "costPerSuccess": [
          0.0060056768444444455,
          0.006987652500000002
        ]
      }
    },
    {
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "attempts": 30,
      "successes": 30,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 0.16981619599999997,
      "costPerAttempt": 0.0056605398666666655,
      "costPerSuccess": 0.0056605398666666655,
      "successfulSpend": 0.16981619599999997,
      "failedSpend": 0,
      "latencyMs": 11406.733333333334,
      "inputTokens": 73381,
      "outputTokens": 12772,
      "reasoningTokens": 4785,
      "cacheReadTokens": 77899,
      "cacheWriteTokens": 0,
      "toolCalls": 116,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.005145948133333333,
          0.0061882832
        ]
      }
    },
    {
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "attempts": 30,
      "successes": 29,
      "successRate": 0.9666666666666667,
      "billingComplete": true,
      "totalCost": 0.178395712,
      "costPerAttempt": 0.005946523733333334,
      "costPerSuccess": 0.006151576275862069,
      "successfulSpend": 0.174233004,
      "failedSpend": 0.004162708,
      "latencyMs": 12450.966666666667,
      "inputTokens": 79660,
      "outputTokens": 13792,
      "reasoningTokens": 3822,
      "cacheReadTokens": 79388,
      "cacheWriteTokens": 0,
      "toolCalls": 149,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          0.9,
          1
        ],
        "costPerSuccess": [
          0.005429113066666669,
          0.006927162814814815
        ]
      }
    },
    {
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "attempts": 30,
      "successes": 30,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 0.22646192799999998,
      "costPerAttempt": 0.0075487309333333325,
      "costPerSuccess": 0.0075487309333333325,
      "successfulSpend": 0.22646192799999998,
      "failedSpend": 0,
      "latencyMs": 14478.833333333334,
      "inputTokens": 94511,
      "outputTokens": 16138,
      "reasoningTokens": 8461,
      "cacheReadTokens": 97622,
      "cacheWriteTokens": 0,
      "toolCalls": 177,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.006567203866666666,
          0.008683957333333332
        ]
      }
    },
    {
      "model": "alibaba/qwen3.8-flash",
      "workflow": "overall",
      "attempts": 90,
      "successes": 90,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 0.07538413199999999,
      "costPerAttempt": 0.0008376014666666666,
      "costPerSuccess": 0.0008376014666666666,
      "successfulSpend": 0.07538413199999999,
      "failedSpend": 0,
      "latencyMs": 14932.322222222223,
      "inputTokens": 2178,
      "outputTokens": 57801,
      "reasoningTokens": 28317,
      "cacheReadTokens": 485687,
      "cacheWriteTokens": 133946,
      "toolCalls": 474,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.0007725256666666672,
          0.0009100238222222223
        ]
      }
    },
    {
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "attempts": 30,
      "successes": 30,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 0.021250084,
      "costPerAttempt": 0.0007083361333333333,
      "costPerSuccess": 0.0007083361333333333,
      "successfulSpend": 0.021250084,
      "failedSpend": 0,
      "latencyMs": 13861.7,
      "inputTokens": 642,
      "outputTokens": 16569,
      "reasoningTokens": 8073,
      "cacheReadTokens": 130514,
      "cacheWriteTokens": 37387,
      "toolCalls": 116,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.0006078264666666665,
          0.0008041272666666666
        ]
      }
    },
    {
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "attempts": 30,
      "successes": 30,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 0.02212009000000001,
      "costPerAttempt": 0.0007373363333333336,
      "costPerSuccess": 0.0007373363333333336,
      "successfulSpend": 0.02212009000000001,
      "failedSpend": 0,
      "latencyMs": 12911.6,
      "inputTokens": 696,
      "outputTokens": 18142,
      "reasoningTokens": 7565,
      "cacheReadTokens": 141715,
      "cacheWriteTokens": 38295,
      "toolCalls": 166,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.0006517104,
          0.0008297728666666667
        ]
      }
    },
    {
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "attempts": 30,
      "successes": 30,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 0.032013958,
      "costPerAttempt": 0.0010671319333333334,
      "costPerSuccess": 0.0010671319333333334,
      "successfulSpend": 0.032013958,
      "failedSpend": 0,
      "latencyMs": 18023.666666666668,
      "inputTokens": 840,
      "outputTokens": 23090,
      "reasoningTokens": 12679,
      "cacheReadTokens": 213458,
      "cacheWriteTokens": 58264,
      "toolCalls": 192,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.0009104941999999999,
          0.0012314424000000003
        ]
      }
    },
    {
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "overall",
      "attempts": 90,
      "successes": 88,
      "successRate": 0.9777777777777777,
      "billingComplete": true,
      "totalCost": 1.09892398,
      "costPerAttempt": 0.012210266444444444,
      "costPerSuccess": 0.0124877725,
      "successfulSpend": 1.0862603099999997,
      "failedSpend": 0.012663669999999998,
      "latencyMs": 34207.066666666666,
      "inputTokens": 2304,
      "outputTokens": 71843,
      "reasoningTokens": 33912,
      "cacheReadTokens": 537594,
      "cacheWriteTokens": 147358,
      "toolCalls": 588,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          0.9444444444444444,
          1
        ],
        "costPerSuccess": [
          0.011590987931034486,
          0.013433031149425283
        ]
      }
    },
    {
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "attempts": 30,
      "successes": 28,
      "successRate": 0.9333333333333333,
      "billingComplete": true,
      "totalCost": 0.3082617199999999,
      "costPerAttempt": 0.010275390666666663,
      "costPerSuccess": 0.01100934714285714,
      "successfulSpend": 0.29559805,
      "failedSpend": 0.012663669999999998,
      "latencyMs": 28806.033333333333,
      "inputTokens": 720,
      "outputTokens": 20525,
      "reasoningTokens": 9244,
      "cacheReadTokens": 156016,
      "cacheWriteTokens": 40674,
      "toolCalls": 162,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          0.8333333333333334,
          1
        ],
        "costPerSuccess": [
          0.010117240689655171,
          0.011817444285714284
        ]
      }
    },
    {
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "attempts": 30,
      "successes": 30,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 0.34329098999999985,
      "costPerAttempt": 0.011443032999999995,
      "costPerSuccess": 0.011443032999999995,
      "successfulSpend": 0.34329098999999985,
      "failedSpend": 0,
      "latencyMs": 33479.36666666667,
      "inputTokens": 780,
      "outputTokens": 23328,
      "reasoningTokens": 10575,
      "cacheReadTokens": 172047,
      "cacheWriteTokens": 43626,
      "toolCalls": 202,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.010082090999999998,
          0.012840167333333328
        ]
      }
    },
    {
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "attempts": 30,
      "successes": 30,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 0.44737127000000004,
      "costPerAttempt": 0.014912375666666668,
      "costPerSuccess": 0.014912375666666668,
      "successfulSpend": 0.44737127000000004,
      "failedSpend": 0,
      "latencyMs": 40335.8,
      "inputTokens": 804,
      "outputTokens": 27990,
      "reasoningTokens": 14093,
      "cacheReadTokens": 209531,
      "cacheWriteTokens": 63058,
      "toolCalls": 224,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.012900292666666665,
          0.01711103133333334
        ]
      }
    },
    {
      "model": "google/gemini-3.8-flash",
      "workflow": "overall",
      "attempts": 90,
      "successes": 90,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 1.01018625,
      "costPerAttempt": 0.011224291666666667,
      "costPerSuccess": 0.011224291666666667,
      "successfulSpend": 1.01018625,
      "failedSpend": 0,
      "latencyMs": 17750.544444444444,
      "inputTokens": 601765,
      "outputTokens": 30156,
      "reasoningTokens": 118874,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0,
      "toolCalls": 480,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.010178233333333335,
          0.012392649999999998
        ]
      }
    },
    {
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "attempts": 30,
      "successes": 30,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 0.29666475000000003,
      "costPerAttempt": 0.009888825,
      "costPerSuccess": 0.009888825,
      "successfulSpend": 0.29666475000000003,
      "failedSpend": 0,
      "latencyMs": 17225.5,
      "inputTokens": 174413,
      "outputTokens": 8198,
      "reasoningTokens": 36030,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0,
      "toolCalls": 129,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.008616,
          0.0112981
        ]
      }
    },
    {
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "attempts": 30,
      "successes": 30,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 0.30820875000000003,
      "costPerAttempt": 0.010273625000000002,
      "costPerSuccess": 0.010273625000000002,
      "successfulSpend": 0.30820875000000003,
      "failedSpend": 0,
      "latencyMs": 16708.366666666665,
      "inputTokens": 178975,
      "outputTokens": 10867,
      "reasoningTokens": 35527,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0,
      "toolCalls": 166,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.008399599999999998,
          0.0129928
        ]
      }
    },
    {
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "attempts": 30,
      "successes": 30,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 0.4053127500000001,
      "costPerAttempt": 0.013510425000000003,
      "costPerSuccess": 0.013510425000000003,
      "successfulSpend": 0.4053127500000001,
      "failedSpend": 0,
      "latencyMs": 19317.766666666666,
      "inputTokens": 248377,
      "outputTokens": 11091,
      "reasoningTokens": 47317,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0,
      "toolCalls": 185,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.011692500000000002,
          0.015688725000000004
        ]
      }
    },
    {
      "model": "openai/gpt-6-astra",
      "workflow": "overall",
      "attempts": 90,
      "successes": 90,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 3.5627145,
      "costPerAttempt": 0.039585716666666666,
      "costPerSuccess": 0.039585716666666666,
      "successfulSpend": 3.5627145,
      "failedSpend": 0,
      "latencyMs": 14148.31111111111,
      "inputTokens": 137742,
      "outputTokens": 27685,
      "reasoningTokens": 0,
      "cacheReadTokens": 210057,
      "cacheWriteTokens": 47279,
      "toolCalls": 392,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.036863266666666665,
          0.04236722222222223
        ]
      }
    },
    {
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "attempts": 30,
      "successes": 30,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 1.1574929999999997,
      "costPerAttempt": 0.03858309999999999,
      "costPerSuccess": 0.03858309999999999,
      "successfulSpend": 1.1574929999999997,
      "failedSpend": 0,
      "latencyMs": 13761.333333333334,
      "inputTokens": 55533,
      "outputTokens": 8444,
      "reasoningTokens": 0,
      "cacheReadTokens": 57763,
      "cacheWriteTokens": 9776,
      "toolCalls": 104,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.0335729,
          0.04388053333333332
        ]
      }
    },
    {
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "attempts": 30,
      "successes": 30,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 1.2438339999999999,
      "costPerAttempt": 0.04146113333333333,
      "costPerSuccess": 0.04146113333333333,
      "successfulSpend": 1.2438339999999999,
      "failedSpend": 0,
      "latencyMs": 13339.466666666667,
      "inputTokens": 59274,
      "outputTokens": 8670,
      "reasoningTokens": 0,
      "cacheReadTokens": 43744,
      "cacheWriteTokens": 13908,
      "toolCalls": 129,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.03721130000000001,
          0.04614103333333333
        ]
      }
    },
    {
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "attempts": 30,
      "successes": 30,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 1.1613874999999998,
      "costPerAttempt": 0.03871291666666666,
      "costPerSuccess": 0.03871291666666666,
      "successfulSpend": 1.1613874999999998,
      "failedSpend": 0,
      "latencyMs": 15344.133333333333,
      "inputTokens": 22935,
      "outputTokens": 10571,
      "reasoningTokens": 0,
      "cacheReadTokens": 108550,
      "cacheWriteTokens": 23595,
      "toolCalls": 159,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.035046799999999996,
          0.04333861666666668
        ]
      }
    },
    {
      "model": "zai/glm-5.3-flash",
      "workflow": "overall",
      "attempts": 90,
      "successes": 89,
      "successRate": 0.9888888888888889,
      "billingComplete": true,
      "totalCost": 0.06913642000000002,
      "costPerAttempt": 0.0007681824444444446,
      "costPerSuccess": 0.0007768137078651687,
      "successfulSpend": 0.06829746000000002,
      "failedSpend": 0.0008389599999999999,
      "latencyMs": 11582.166666666666,
      "inputTokens": 323990,
      "outputTokens": 32854,
      "reasoningTokens": 5242,
      "cacheReadTokens": 49664,
      "cacheWriteTokens": 0,
      "toolCalls": 401,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          0.9666666666666667,
          1
        ],
        "costPerSuccess": [
          0.0007317433333333332,
          0.0008253292134831461
        ]
      }
    },
    {
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "attempts": 30,
      "successes": 30,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 0.020482339999999998,
      "costPerAttempt": 0.0006827446666666666,
      "costPerSuccess": 0.0006827446666666666,
      "successfulSpend": 0.020482339999999998,
      "failedSpend": 0,
      "latencyMs": 8911.566666666668,
      "inputTokens": 96204,
      "outputTokens": 9518,
      "reasoningTokens": 1541,
      "cacheReadTokens": 17408,
      "cacheWriteTokens": 0,
      "toolCalls": 100,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.0006309606666666665,
          0.0007391570000000002
        ]
      }
    },
    {
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "attempts": 30,
      "successes": 30,
      "successRate": 1,
      "billingComplete": true,
      "totalCost": 0.020955530000000003,
      "costPerAttempt": 0.0006985176666666668,
      "costPerSuccess": 0.0006985176666666668,
      "successfulSpend": 0.020955530000000003,
      "failedSpend": 0,
      "latencyMs": 10490.933333333332,
      "inputTokens": 102421,
      "outputTokens": 9071,
      "reasoningTokens": 1100,
      "cacheReadTokens": 16896,
      "cacheWriteTokens": 0,
      "toolCalls": 142,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          1,
          1
        ],
        "costPerSuccess": [
          0.0006323776666666667,
          0.000768733
        ]
      }
    },
    {
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "attempts": 30,
      "successes": 29,
      "successRate": 0.9666666666666667,
      "billingComplete": true,
      "totalCost": 0.027698549999999995,
      "costPerAttempt": 0.0009232849999999998,
      "costPerSuccess": 0.0009551224137931033,
      "successfulSpend": 0.026859589999999996,
      "failedSpend": 0.0008389599999999999,
      "latencyMs": 15344,
      "inputTokens": 125365,
      "outputTokens": 14265,
      "reasoningTokens": 2601,
      "cacheReadTokens": 15360,
      "cacheWriteTokens": 0,
      "toolCalls": 159,
      "unresolvedRuns": 0,
      "ci": {
        "method": "percentile whole-task bootstrap, stratified by workflow",
        "seed": 90826,
        "iterations": 2000,
        "successRate": [
          0.9,
          1
        ],
        "costPerSuccess": [
          0.0008468796551724135,
          0.001073276551724138
        ]
      }
    }
  ],
  "overall": {
    "attempts": 990,
    "successes": 979,
    "successRate": 0.9888888888888889,
    "billingComplete": true,
    "totalCost": 11.82080844599999,
    "costPerAttempt": 0.011940210551515142,
    "costPerSuccess": 0.012074370220633289,
    "successfulSpend": 11.673367164999995,
    "failedSpend": 0.147441281,
    "latencyMs": 14233.470707070706,
    "inputTokens": 1658942,
    "outputTokens": 506394,
    "reasoningTokens": 232633,
    "cacheReadTokens": 3307514,
    "cacheWriteTokens": 955838,
    "toolCalls": 4834,
    "unresolvedRuns": 0
  },
  "rows": [
    {
      "trialId": "scored:rec-001:openai/gpt-5.6-sol:0",
      "fixtureId": "rec-001",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.009016,
      "latencyMs": 7174,
      "toolCalls": 3,
      "inputTokens": 3283,
      "outputTokens": 211,
      "reasoningTokens": 34,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-001:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "rec-001",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000186714,
      "latencyMs": 4202,
      "toolCalls": 3,
      "inputTokens": 840,
      "outputTokens": 522,
      "reasoningTokens": 0,
      "cacheReadTokens": 3072,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-001:openai/gpt-5.6-luna:0",
      "fixtureId": "rec-001",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0010172000000000002,
      "latencyMs": 8094,
      "toolCalls": 3,
      "inputTokens": 3286,
      "outputTokens": 236,
      "reasoningTokens": 64,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-001:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "rec-001",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01021777,
      "latencyMs": 29257,
      "toolCalls": 6,
      "inputTokens": 30,
      "outputTokens": 685,
      "reasoningTokens": 289,
      "cacheReadTokens": 6081,
      "cacheWriteTokens": 1312
    },
    {
      "trialId": "scored:rec-001:google/gemini-3.8-flash:0",
      "fixtureId": "rec-001",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0071775,
      "latencyMs": 11654,
      "toolCalls": 4,
      "inputTokens": 4180,
      "outputTokens": 267,
      "reasoningTokens": 811,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-001:anthropic/claude-sonnet-5:0",
      "fixtureId": "rec-001",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0100743,
      "latencyMs": 7685,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 382,
      "reasoningTokens": 0,
      "cacheReadTokens": 3624,
      "cacheWriteTokens": 2207
    },
    {
      "trialId": "scored:rec-001:alibaba/qwen3.8-flash:0",
      "fixtureId": "rec-001",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000663462,
      "latencyMs": 11438,
      "toolCalls": 5,
      "inputTokens": 30,
      "outputTokens": 616,
      "reasoningTokens": 153,
      "cacheReadTokens": 5802,
      "cacheWriteTokens": 1022
    },
    {
      "trialId": "scored:rec-001:anthropic/claude-opus-5:0",
      "fixtureId": "rec-001",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.021167,
      "latencyMs": 8143,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 403,
      "reasoningTokens": 31,
      "cacheReadTokens": 4349,
      "cacheWriteTokens": 1298
    },
    {
      "trialId": "scored:rec-001:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "rec-001",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0036340919999999994,
      "latencyMs": 6472,
      "toolCalls": 3,
      "inputTokens": 1452,
      "outputTokens": 329,
      "reasoningTokens": 80,
      "cacheReadTokens": 2223,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-003:openai/gpt-5.6-sol:0",
      "fixtureId": "rec-003",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0108124,
      "latencyMs": 11629,
      "toolCalls": 7,
      "inputTokens": 759,
      "outputTokens": 434,
      "reasoningTokens": 58,
      "cacheReadTokens": 4722,
      "cacheWriteTokens": 1372
    },
    {
      "trialId": "scored:rec-003:openai/gpt-5.6-luna:0",
      "fixtureId": "rec-003",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00109297,
      "latencyMs": 7949,
      "toolCalls": 5,
      "inputTokens": 753,
      "outputTokens": 384,
      "reasoningTokens": 94,
      "cacheReadTokens": 2326,
      "cacheWriteTokens": 1289
    },
    {
      "trialId": "scored:rec-003:anthropic/claude-opus-5:0",
      "fixtureId": "rec-003",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0366595,
      "latencyMs": 11584,
      "toolCalls": 6,
      "inputTokens": 8,
      "outputTokens": 780,
      "reasoningTokens": 33,
      "cacheReadTokens": 7589,
      "cacheWriteTokens": 2000
    },
    {
      "trialId": "scored:rec-003:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "rec-003",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000298549,
      "latencyMs": 6460,
      "toolCalls": 7,
      "inputTokens": 1437,
      "outputTokens": 769,
      "reasoningTokens": 0,
      "cacheReadTokens": 5120,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-003:alibaba/qwen3.8-flash:0",
      "fixtureId": "rec-003",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0006401600000000001,
      "latencyMs": 8457,
      "toolCalls": 6,
      "inputTokens": 18,
      "outputTokens": 430,
      "reasoningTokens": 226,
      "cacheReadTokens": 3510,
      "cacheWriteTokens": 1364
    },
    {
      "trialId": "scored:rec-003:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "rec-003",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.004767884,
      "latencyMs": 7547,
      "toolCalls": 6,
      "inputTokens": 1954,
      "outputTokens": 420,
      "reasoningTokens": 88,
      "cacheReadTokens": 4021,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-003:google/gemini-3.8-flash:0",
      "fixtureId": "rec-003",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.010337250000000001,
      "latencyMs": 14480,
      "toolCalls": 7,
      "inputTokens": 6818,
      "outputTokens": 557,
      "reasoningTokens": 836,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-003:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "rec-003",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0148358,
      "latencyMs": 36150,
      "toolCalls": 8,
      "inputTokens": 30,
      "outputTokens": 781,
      "reasoningTokens": 671,
      "cacheReadTokens": 7640,
      "cacheWriteTokens": 1906
    },
    {
      "trialId": "scored:rec-003:anthropic/claude-sonnet-5:0",
      "fixtureId": "rec-003",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0167525,
      "latencyMs": 11202,
      "toolCalls": 5,
      "inputTokens": 6,
      "outputTokens": 737,
      "reasoningTokens": 120,
      "cacheReadTokens": 4340,
      "cacheWriteTokens": 2921
    },
    {
      "trialId": "scored:ext-003:anthropic/claude-opus-5:0",
      "fixtureId": "ext-003",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.024982250000000004,
      "latencyMs": 7766,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 409,
      "reasoningTokens": 0,
      "cacheReadTokens": 5517,
      "cacheWriteTokens": 1915
    },
    {
      "trialId": "scored:ext-003:openai/gpt-5.6-luna:0",
      "fixtureId": "ext-003",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0009035100000000001,
      "latencyMs": 6761,
      "toolCalls": 2,
      "inputTokens": 951,
      "outputTokens": 241,
      "reasoningTokens": 57,
      "cacheReadTokens": 2298,
      "cacheWriteTokens": 1239
    },
    {
      "trialId": "scored:ext-003:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "ext-003",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.004443208000000001,
      "latencyMs": 11234,
      "toolCalls": 3,
      "inputTokens": 1864,
      "outputTokens": 372,
      "reasoningTokens": 80,
      "cacheReadTokens": 4382,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-003:google/gemini-3.8-flash:0",
      "fixtureId": "ext-003",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008844,
      "latencyMs": 12677,
      "toolCalls": 3,
      "inputTokens": 5502,
      "outputTokens": 172,
      "reasoningTokens": 1086,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-003:alibaba/qwen3.8-flash:0",
      "fixtureId": "ext-003",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0005077840000000001,
      "latencyMs": 7304,
      "toolCalls": 2,
      "inputTokens": 18,
      "outputTokens": 371,
      "reasoningTokens": 105,
      "cacheReadTokens": 3599,
      "cacheWriteTokens": 1118
    },
    {
      "trialId": "scored:ext-003:anthropic/claude-sonnet-5:0",
      "fixtureId": "ext-003",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0122517,
      "latencyMs": 7814,
      "toolCalls": 2,
      "inputTokens": 6,
      "outputTokens": 434,
      "reasoningTokens": 9,
      "cacheReadTokens": 4711,
      "cacheWriteTokens": 2747
    },
    {
      "trialId": "scored:ext-003:openai/gpt-5.6-sol:0",
      "fixtureId": "ext-003",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0080879,
      "latencyMs": 8052,
      "toolCalls": 2,
      "inputTokens": 951,
      "outputTokens": 241,
      "reasoningTokens": 23,
      "cacheReadTokens": 2292,
      "cacheWriteTokens": 1235
    },
    {
      "trialId": "scored:ext-003:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "ext-003",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": false,
      "reasons": [
        "submission.params.taxId: required property missing",
        "submission.params.streetAddress: required property missing",
        "submission.params.city: required property missing",
        "submission.params.state: required property missing",
        "submission.params.postalCode: required property missing",
        "submission.params.country: required property missing",
        "submission.params.primaryContactName: required property missing",
        "submission.params.primaryContactEmail: required property missing",
        "submission.params.primaryContactPhone: required property missing"
      ],
      "cost": 0.007604629999999999,
      "latencyMs": 21140,
      "toolCalls": 3,
      "inputTokens": 18,
      "outputTokens": 334,
      "reasoningTokens": 325,
      "cacheReadTokens": 3689,
      "cacheWriteTokens": 1195
    },
    {
      "trialId": "scored:ext-003:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "ext-003",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000267011,
      "latencyMs": 13356,
      "toolCalls": 3,
      "inputTokens": 1653,
      "outputTokens": 479,
      "reasoningTokens": 0,
      "cacheReadTokens": 4864,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-004:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "sup-004",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.005536872,
      "latencyMs": 10552,
      "toolCalls": 5,
      "inputTokens": 2089,
      "outputTokens": 534,
      "reasoningTokens": 135,
      "cacheReadTokens": 2958,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-004:openai/gpt-5.6-sol:0",
      "fixtureId": "sup-004",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0089536,
      "latencyMs": 45531,
      "toolCalls": 5,
      "inputTokens": 843,
      "outputTokens": 344,
      "reasoningTokens": 38,
      "cacheReadTokens": 1418,
      "cacheWriteTokens": 1248
    },
    {
      "trialId": "scored:sup-004:google/gemini-3.8-flash:0",
      "fixtureId": "sup-004",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.013336500000000001,
      "latencyMs": 48964,
      "toolCalls": 5,
      "inputTokens": 9227,
      "outputTokens": 387,
      "reasoningTokens": 1324,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-004:anthropic/claude-sonnet-5:0",
      "fixtureId": "sup-004",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0164294,
      "latencyMs": 10937,
      "toolCalls": 5,
      "inputTokens": 6,
      "outputTokens": 615,
      "reasoningTokens": 137,
      "cacheReadTokens": 4737,
      "cacheWriteTokens": 3180
    },
    {
      "trialId": "scored:sup-004:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "sup-004",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000289403,
      "latencyMs": 4872,
      "toolCalls": 5,
      "inputTokens": 1484,
      "outputTokens": 803,
      "reasoningTokens": 0,
      "cacheReadTokens": 3840,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-004:alibaba/qwen3.8-flash:0",
      "fixtureId": "sup-004",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000714564,
      "latencyMs": 10315,
      "toolCalls": 5,
      "inputTokens": 18,
      "outputTokens": 541,
      "reasoningTokens": 193,
      "cacheReadTokens": 3744,
      "cacheWriteTokens": 1534
    },
    {
      "trialId": "scored:sup-004:anthropic/claude-opus-5:0",
      "fixtureId": "sup-004",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0357945,
      "latencyMs": 11870,
      "toolCalls": 5,
      "inputTokens": 6,
      "outputTokens": 682,
      "reasoningTokens": 66,
      "cacheReadTokens": 5454,
      "cacheWriteTokens": 2294
    },
    {
      "trialId": "scored:sup-004:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "sup-004",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.013564529999999998,
      "latencyMs": 32551,
      "toolCalls": 6,
      "inputTokens": 24,
      "outputTokens": 846,
      "reasoningTokens": 467,
      "cacheReadTokens": 5859,
      "cacheWriteTokens": 1857
    },
    {
      "trialId": "scored:sup-004:openai/gpt-5.6-luna:0",
      "fixtureId": "sup-004",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0011963800000000001,
      "latencyMs": 7205,
      "toolCalls": 5,
      "inputTokens": 844,
      "outputTokens": 329,
      "reasoningTokens": 160,
      "cacheReadTokens": 2764,
      "cacheWriteTokens": 1542
    },
    {
      "trialId": "scored:ext-007:openai/gpt-5.6-sol:0",
      "fixtureId": "ext-007",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0084294,
      "latencyMs": 8094,
      "toolCalls": 4,
      "inputTokens": 837,
      "outputTokens": 282,
      "reasoningTokens": 42,
      "cacheReadTokens": 2252,
      "cacheWriteTokens": 1226
    },
    {
      "trialId": "scored:ext-007:alibaba/qwen3.8-flash:0",
      "fixtureId": "ext-007",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000631768,
      "latencyMs": 9062,
      "toolCalls": 4,
      "inputTokens": 18,
      "outputTokens": 558,
      "reasoningTokens": 146,
      "cacheReadTokens": 3463,
      "cacheWriteTokens": 1213
    },
    {
      "trialId": "scored:ext-007:anthropic/claude-sonnet-5:0",
      "fixtureId": "ext-007",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.014081100000000001,
      "latencyMs": 10403,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 548,
      "reasoningTokens": 76,
      "cacheReadTokens": 4458,
      "cacheWriteTokens": 2775
    },
    {
      "trialId": "scored:ext-007:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "ext-007",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.004671964,
      "latencyMs": 7266,
      "toolCalls": 5,
      "inputTokens": 1851,
      "outputTokens": 436,
      "reasoningTokens": 81,
      "cacheReadTokens": 4121,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-007:anthropic/claude-opus-5:0",
      "fixtureId": "ext-007",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.035844249999999994,
      "latencyMs": 15091,
      "toolCalls": 6,
      "inputTokens": 8,
      "outputTokens": 735,
      "reasoningTokens": 21,
      "cacheReadTokens": 7946,
      "cacheWriteTokens": 2069
    },
    {
      "trialId": "scored:ext-007:google/gemini-3.8-flash:0",
      "fixtureId": "ext-007",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.011057250000000001,
      "latencyMs": 19147,
      "toolCalls": 6,
      "inputTokens": 7178,
      "outputTokens": 267,
      "reasoningTokens": 1246,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-007:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "ext-007",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00999539,
      "latencyMs": 25107,
      "toolCalls": 6,
      "inputTokens": 24,
      "outputTokens": 622,
      "reasoningTokens": 289,
      "cacheReadTokens": 5317,
      "cacheWriteTokens": 1431
    },
    {
      "trialId": "scored:ext-007:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "ext-007",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00030937900000000003,
      "latencyMs": 6425,
      "toolCalls": 5,
      "inputTokens": 2207,
      "outputTokens": 551,
      "reasoningTokens": 0,
      "cacheReadTokens": 4096,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-007:openai/gpt-5.6-luna:0",
      "fixtureId": "ext-007",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00096923,
      "latencyMs": 7400,
      "toolCalls": 4,
      "inputTokens": 837,
      "outputTokens": 282,
      "reasoningTokens": 89,
      "cacheReadTokens": 2269,
      "cacheWriteTokens": 1245
    },
    {
      "trialId": "scored:sup-002:openai/gpt-5.6-luna:0",
      "fixtureId": "sup-002",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00098486,
      "latencyMs": 7768,
      "toolCalls": 4,
      "inputTokens": 803,
      "outputTokens": 311,
      "reasoningTokens": 72,
      "cacheReadTokens": 2333,
      "cacheWriteTokens": 1272
    },
    {
      "trialId": "scored:sup-002:anthropic/claude-opus-5:0",
      "fixtureId": "sup-002",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0382925,
      "latencyMs": 12375,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 830,
      "reasoningTokens": 75,
      "cacheReadTokens": 5125,
      "cacheWriteTokens": 2092
    },
    {
      "trialId": "scored:sup-002:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "sup-002",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.004420812,
      "latencyMs": 10445,
      "toolCalls": 4,
      "inputTokens": 1827,
      "outputTokens": 386,
      "reasoningTokens": 91,
      "cacheReadTokens": 2733,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-002:alibaba/qwen3.8-flash:0",
      "fixtureId": "sup-002",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00067869,
      "latencyMs": 10195,
      "toolCalls": 4,
      "inputTokens": 18,
      "outputTokens": 584,
      "reasoningTokens": 159,
      "cacheReadTokens": 3525,
      "cacheWriteTokens": 1351
    },
    {
      "trialId": "scored:sup-002:google/gemini-3.8-flash:0",
      "fixtureId": "sup-002",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.009272250000000001,
      "latencyMs": 13234,
      "toolCalls": 5,
      "inputTokens": 5923,
      "outputTokens": 352,
      "reasoningTokens": 936,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-002:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "sup-002",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000277722,
      "latencyMs": 6272,
      "toolCalls": 5,
      "inputTokens": 1651,
      "outputTokens": 550,
      "reasoningTokens": 0,
      "cacheReadTokens": 4864,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-002:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "sup-002",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01100605,
      "latencyMs": 26937,
      "toolCalls": 7,
      "inputTokens": 24,
      "outputTokens": 816,
      "reasoningTokens": 194,
      "cacheReadTokens": 5415,
      "cacheWriteTokens": 1591
    },
    {
      "trialId": "scored:sup-002:openai/gpt-5.6-sol:0",
      "fixtureId": "sup-002",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0084744,
      "latencyMs": 8608,
      "toolCalls": 4,
      "inputTokens": 803,
      "outputTokens": 284,
      "reasoningTokens": 42,
      "cacheReadTokens": 2317,
      "cacheWriteTokens": 1258
    },
    {
      "trialId": "scored:rec-002:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "rec-002",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0045294919999999995,
      "latencyMs": 7742,
      "toolCalls": 4,
      "inputTokens": 1703,
      "outputTokens": 390,
      "reasoningTokens": 159,
      "cacheReadTokens": 2443,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-002:anthropic/claude-sonnet-5:0",
      "fixtureId": "sup-002",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.014928199999999999,
      "latencyMs": 13337,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 691,
      "reasoningTokens": 9,
      "cacheReadTokens": 4406,
      "cacheWriteTokens": 2814
    },
    {
      "trialId": "scored:rec-002:openai/gpt-5.6-sol:0",
      "fixtureId": "rec-002",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0090067,
      "latencyMs": 9140,
      "toolCalls": 4,
      "inputTokens": 1602,
      "outputTokens": 261,
      "reasoningTokens": 31,
      "cacheReadTokens": 1051,
      "cacheWriteTokens": 1069
    },
    {
      "trialId": "scored:rec-002:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "rec-002",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000278862,
      "latencyMs": 5504,
      "toolCalls": 5,
      "inputTokens": 1911,
      "outputTokens": 522,
      "reasoningTokens": 0,
      "cacheReadTokens": 3840,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-002:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "rec-002",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.010729820000000001,
      "latencyMs": 32938,
      "toolCalls": 6,
      "inputTokens": 24,
      "outputTokens": 875,
      "reasoningTokens": 261,
      "cacheReadTokens": 5196,
      "cacheWriteTokens": 1193
    },
    {
      "trialId": "scored:rec-002:anthropic/claude-sonnet-5:0",
      "fixtureId": "rec-002",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0121138,
      "latencyMs": 10071,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 449,
      "reasoningTokens": 56,
      "cacheReadTokens": 3984,
      "cacheWriteTokens": 2502
    },
    {
      "trialId": "scored:rec-002:alibaba/qwen3.8-flash:0",
      "fixtureId": "rec-002",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000642918,
      "latencyMs": 11572,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 554,
      "reasoningTokens": 171,
      "cacheReadTokens": 4833,
      "cacheWriteTokens": 1105
    },
    {
      "trialId": "scored:rec-002:anthropic/claude-opus-5:0",
      "fixtureId": "rec-002",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.025167500000000002,
      "latencyMs": 8961,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 497,
      "reasoningTokens": 25,
      "cacheReadTokens": 4700,
      "cacheWriteTokens": 1558
    },
    {
      "trialId": "scored:rec-002:google/gemini-3.8-flash:0",
      "fixtureId": "rec-002",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0079125,
      "latencyMs": 13894,
      "toolCalls": 5,
      "inputTokens": 4920,
      "outputTokens": 364,
      "reasoningTokens": 762,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-002:openai/gpt-5.6-luna:0",
      "fixtureId": "rec-002",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00100286,
      "latencyMs": 6736,
      "toolCalls": 4,
      "inputTokens": 1606,
      "outputTokens": 258,
      "reasoningTokens": 66,
      "cacheReadTokens": 1068,
      "cacheWriteTokens": 1086
    },
    {
      "trialId": "scored:ext-010:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "ext-010",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.005844388,
      "latencyMs": 11554,
      "toolCalls": 5,
      "inputTokens": 1920,
      "outputTokens": 495,
      "reasoningTokens": 295,
      "cacheReadTokens": 4127,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-010:alibaba/qwen3.8-flash:0",
      "fixtureId": "ext-010",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0008326519999999999,
      "latencyMs": 15341,
      "toolCalls": 4,
      "inputTokens": 24,
      "outputTokens": 619,
      "reasoningTokens": 403,
      "cacheReadTokens": 4917,
      "cacheWriteTokens": 1349
    },
    {
      "trialId": "scored:ext-010:google/gemini-3.8-flash:0",
      "fixtureId": "ext-010",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.015405,
      "latencyMs": 24858,
      "toolCalls": 5,
      "inputTokens": 9825,
      "outputTokens": 249,
      "reasoningTokens": 1894,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-010:anthropic/claude-opus-5:0",
      "fixtureId": "ext-010",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0346155,
      "latencyMs": 15122,
      "toolCalls": 4,
      "inputTokens": 8,
      "outputTokens": 672,
      "reasoningTokens": 155,
      "cacheReadTokens": 6801,
      "cacheWriteTokens": 1680
    },
    {
      "trialId": "scored:ext-010:openai/gpt-5.6-sol:0",
      "fixtureId": "ext-010",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0095394,
      "latencyMs": 9720,
      "toolCalls": 3,
      "inputTokens": 1545,
      "outputTokens": 283,
      "reasoningTokens": 68,
      "cacheReadTokens": 1072,
      "cacheWriteTokens": 1090
    },
    {
      "trialId": "scored:ext-010:anthropic/claude-sonnet-5:0",
      "fixtureId": "ext-010",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0195954,
      "latencyMs": 14815,
      "toolCalls": 4,
      "inputTokens": 8,
      "outputTokens": 658,
      "reasoningTokens": 451,
      "cacheReadTokens": 6422,
      "cacheWriteTokens": 2882
    },
    {
      "trialId": "scored:ext-010:openai/gpt-5.6-luna:0",
      "fixtureId": "ext-010",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": false,
      "reasons": [
        "params.totalCents: substantive value or membership differs"
      ],
      "cost": 0.0010566999999999998,
      "latencyMs": 9380,
      "toolCalls": 3,
      "inputTokens": 1547,
      "outputTokens": 251,
      "reasoningTokens": 116,
      "cacheReadTokens": 1120,
      "cacheWriteTokens": 1138
    },
    {
      "trialId": "scored:ext-010:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "ext-010",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000307285,
      "latencyMs": 5506,
      "toolCalls": 4,
      "inputTokens": 1629,
      "outputTokens": 801,
      "reasoningTokens": 0,
      "cacheReadTokens": 4352,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-010:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "ext-010",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01276675,
      "latencyMs": 34898,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 788,
      "reasoningTokens": 538,
      "cacheReadTokens": 5075,
      "cacheWriteTokens": 1560
    },
    {
      "trialId": "scored:ext-001:anthropic/claude-sonnet-5:0",
      "fixtureId": "ext-001",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0138337,
      "latencyMs": 11400,
      "toolCalls": 2,
      "inputTokens": 6,
      "outputTokens": 600,
      "reasoningTokens": 0,
      "cacheReadTokens": 4521,
      "cacheWriteTokens": 2767
    },
    {
      "trialId": "scored:ext-001:anthropic/claude-opus-5:0",
      "fixtureId": "ext-001",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.027125749999999997,
      "latencyMs": 8209,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 493,
      "reasoningTokens": 0,
      "cacheReadTokens": 5329,
      "cacheWriteTokens": 1937
    },
    {
      "trialId": "scored:ext-001:alibaba/qwen3.8-flash:0",
      "fixtureId": "ext-001",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000907684,
      "latencyMs": 14551,
      "toolCalls": 3,
      "inputTokens": 24,
      "outputTokens": 551,
      "reasoningTokens": 499,
      "cacheReadTokens": 5659,
      "cacheWriteTokens": 1599
    },
    {
      "trialId": "scored:ext-001:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "ext-001",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.006165016000000001,
      "latencyMs": 13358,
      "toolCalls": 3,
      "inputTokens": 2131,
      "outputTokens": 546,
      "reasoningTokens": 253,
      "cacheReadTokens": 4274,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-001:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "ext-001",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000363715,
      "latencyMs": 7440,
      "toolCalls": 7,
      "inputTokens": 1759,
      "outputTokens": 871,
      "reasoningTokens": 0,
      "cacheReadTokens": 6912,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-001:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "ext-001",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.010790319999999999,
      "latencyMs": 29343,
      "toolCalls": 6,
      "inputTokens": 24,
      "outputTokens": 776,
      "reasoningTokens": 283,
      "cacheReadTokens": 6196,
      "cacheWriteTokens": 1334
    },
    {
      "trialId": "scored:ext-001:google/gemini-3.8-flash:0",
      "fixtureId": "ext-001",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00898875,
      "latencyMs": 14629,
      "toolCalls": 2,
      "inputTokens": 4710,
      "outputTokens": 239,
      "reasoningTokens": 1216,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-001:openai/gpt-5.6-luna:0",
      "fixtureId": "ext-001",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00105481,
      "latencyMs": 7520,
      "toolCalls": 2,
      "inputTokens": 883,
      "outputTokens": 325,
      "reasoningTokens": 85,
      "cacheReadTokens": 2448,
      "cacheWriteTokens": 1349
    },
    {
      "trialId": "scored:ext-001:openai/gpt-5.6-sol:0",
      "fixtureId": "ext-001",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0089165,
      "latencyMs": 9183,
      "toolCalls": 2,
      "inputTokens": 883,
      "outputTokens": 319,
      "reasoningTokens": 19,
      "cacheReadTokens": 2415,
      "cacheWriteTokens": 1315
    },
    {
      "trialId": "scored:rec-010:openai/gpt-5.6-sol:0",
      "fixtureId": "rec-010",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.009694,
      "latencyMs": 8207,
      "toolCalls": 6,
      "inputTokens": 1659,
      "outputTokens": 293,
      "reasoningTokens": 35,
      "cacheReadTokens": 1130,
      "cacheWriteTokens": 1148
    },
    {
      "trialId": "scored:rec-010:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "rec-010",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.005479804,
      "latencyMs": 10725,
      "toolCalls": 7,
      "inputTokens": 1949,
      "outputTokens": 488,
      "reasoningTokens": 204,
      "cacheReadTokens": 3791,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-010:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "rec-010",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000264229,
      "latencyMs": 5178,
      "toolCalls": 7,
      "inputTokens": 1250,
      "outputTokens": 661,
      "reasoningTokens": 0,
      "cacheReadTokens": 4864,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-010:alibaba/qwen3.8-flash:0",
      "fixtureId": "rec-010",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000918374,
      "latencyMs": 17174,
      "toolCalls": 9,
      "inputTokens": 30,
      "outputTokens": 678,
      "reasoningTokens": 375,
      "cacheReadTokens": 7004,
      "cacheWriteTokens": 1533
    },
    {
      "trialId": "scored:rec-010:google/gemini-3.8-flash:0",
      "fixtureId": "rec-010",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.009165,
      "latencyMs": 14456,
      "toolCalls": 7,
      "inputTokens": 5695,
      "outputTokens": 260,
      "reasoningTokens": 1045,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-010:anthropic/claude-opus-5:0",
      "fixtureId": "rec-010",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.03742625,
      "latencyMs": 12955,
      "toolCalls": 8,
      "inputTokens": 8,
      "outputTokens": 783,
      "reasoningTokens": 51,
      "cacheReadTokens": 7685,
      "cacheWriteTokens": 2031
    },
    {
      "trialId": "scored:rec-010:openai/gpt-5.6-luna:0",
      "fixtureId": "rec-010",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0013377,
      "latencyMs": 6906,
      "toolCalls": 6,
      "inputTokens": 1662,
      "outputTokens": 271,
      "reasoningTokens": 83,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 2322
    },
    {
      "trialId": "scored:rec-010:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "rec-010",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01273958,
      "latencyMs": 34726,
      "toolCalls": 8,
      "inputTokens": 24,
      "outputTokens": 810,
      "reasoningTokens": 447,
      "cacheReadTokens": 5174,
      "cacheWriteTokens": 1708
    },
    {
      "trialId": "scored:rec-010:anthropic/claude-sonnet-5:0",
      "fixtureId": "rec-010",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0161859,
      "latencyMs": 10693,
      "toolCalls": 6,
      "inputTokens": 6,
      "outputTokens": 728,
      "reasoningTokens": 112,
      "cacheReadTokens": 4157,
      "cacheWriteTokens": 2777
    },
    {
      "trialId": "scored:sup-010:alibaba/qwen3.8-flash:0",
      "fixtureId": "sup-010",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000799992,
      "latencyMs": 12038,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 607,
      "reasoningTokens": 241,
      "cacheReadTokens": 5262,
      "cacheWriteTokens": 1567
    },
    {
      "trialId": "scored:sup-010:openai/gpt-5.6-sol:0",
      "fixtureId": "sup-010",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0083262,
      "latencyMs": 6067,
      "toolCalls": 4,
      "inputTokens": 901,
      "outputTokens": 217,
      "reasoningTokens": 51,
      "cacheReadTokens": 1246,
      "cacheWriteTokens": 1438
    },
    {
      "trialId": "scored:sup-010:anthropic/claude-sonnet-5:0",
      "fixtureId": "sup-010",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.018439,
      "latencyMs": 14552,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 836,
      "reasoningTokens": 85,
      "cacheReadTokens": 4860,
      "cacheWriteTokens": 3298
    },
    {
      "trialId": "scored:sup-010:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "sup-010",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000301737,
      "latencyMs": 3660,
      "toolCalls": 4,
      "inputTokens": 1954,
      "outputTokens": 697,
      "reasoningTokens": 0,
      "cacheReadTokens": 3328,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-010:anthropic/claude-opus-5:0",
      "fixtureId": "sup-010",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0426945,
      "latencyMs": 14737,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 961,
      "reasoningTokens": 34,
      "cacheReadTokens": 5579,
      "cacheWriteTokens": 2400
    },
    {
      "trialId": "scored:sup-010:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "sup-010",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.005930276,
      "latencyMs": 12404,
      "toolCalls": 5,
      "inputTokens": 2209,
      "outputTokens": 546,
      "reasoningTokens": 181,
      "cacheReadTokens": 3079,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-010:openai/gpt-5.6-luna:0",
      "fixtureId": "sup-010",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00103695,
      "latencyMs": 8272,
      "toolCalls": 4,
      "inputTokens": 904,
      "outputTokens": 299,
      "reasoningTokens": 68,
      "cacheReadTokens": 2675,
      "cacheWriteTokens": 1449
    },
    {
      "trialId": "scored:sup-010:google/gemini-3.8-flash:0",
      "fixtureId": "sup-010",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0147375,
      "latencyMs": 23781,
      "toolCalls": 4,
      "inputTokens": 7850,
      "outputTokens": 372,
      "reasoningTokens": 1988,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-001:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "sup-001",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000281679,
      "latencyMs": 5324,
      "toolCalls": 5,
      "inputTokens": 1173,
      "outputTokens": 907,
      "reasoningTokens": 0,
      "cacheReadTokens": 3840,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-010:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "sup-010",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01175799,
      "latencyMs": 30686,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 807,
      "reasoningTokens": 284,
      "cacheReadTokens": 5347,
      "cacheWriteTokens": 1702
    },
    {
      "trialId": "scored:sup-001:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "sup-001",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.005751592,
      "latencyMs": 11769,
      "toolCalls": 6,
      "inputTokens": 2097,
      "outputTokens": 527,
      "reasoningTokens": 181,
      "cacheReadTokens": 4088,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-001:openai/gpt-5.6-luna:0",
      "fixtureId": "sup-001",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0010065500000000002,
      "latencyMs": 7796,
      "toolCalls": 5,
      "inputTokens": 750,
      "outputTokens": 307,
      "reasoningTokens": 87,
      "cacheReadTokens": 2450,
      "cacheWriteTokens": 1339
    },
    {
      "trialId": "scored:sup-001:anthropic/claude-sonnet-5:0",
      "fixtureId": "sup-001",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0157641,
      "latencyMs": 13462,
      "toolCalls": 5,
      "inputTokens": 6,
      "outputTokens": 619,
      "reasoningTokens": 126,
      "cacheReadTokens": 4398,
      "cacheWriteTokens": 2969
    },
    {
      "trialId": "scored:sup-001:google/gemini-3.8-flash:0",
      "fixtureId": "sup-001",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.007169999999999999,
      "latencyMs": 12196,
      "toolCalls": 5,
      "inputTokens": 4110,
      "outputTokens": 285,
      "reasoningTokens": 805,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-001:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "sup-001",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01138003,
      "latencyMs": 29574,
      "toolCalls": 7,
      "inputTokens": 24,
      "outputTokens": 800,
      "reasoningTokens": 248,
      "cacheReadTokens": 5509,
      "cacheWriteTokens": 1643
    },
    {
      "trialId": "scored:sup-001:anthropic/claude-opus-5:0",
      "fixtureId": "sup-001",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.03451975,
      "latencyMs": 11797,
      "toolCalls": 5,
      "inputTokens": 6,
      "outputTokens": 681,
      "reasoningTokens": 68,
      "cacheReadTokens": 5117,
      "cacheWriteTokens": 2113
    },
    {
      "trialId": "scored:sup-001:openai/gpt-5.6-sol:0",
      "fixtureId": "sup-001",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0089534,
      "latencyMs": 9417,
      "toolCalls": 5,
      "inputTokens": 750,
      "outputTokens": 313,
      "reasoningTokens": 51,
      "cacheReadTokens": 2442,
      "cacheWriteTokens": 1330
    },
    {
      "trialId": "scored:sup-001:alibaba/qwen3.8-flash:0",
      "fixtureId": "sup-001",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0010465460000000002,
      "latencyMs": 16105,
      "toolCalls": 7,
      "inputTokens": 30,
      "outputTokens": 797,
      "reasoningTokens": 370,
      "cacheReadTokens": 7091,
      "cacheWriteTokens": 1899
    },
    {
      "trialId": "scored:rec-005:anthropic/claude-sonnet-5:0",
      "fixtureId": "rec-005",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0101186,
      "latencyMs": 9725,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 357,
      "reasoningTokens": 50,
      "cacheReadTokens": 3508,
      "cacheWriteTokens": 2134
    },
    {
      "trialId": "scored:rec-005:alibaba/qwen3.8-flash:0",
      "fixtureId": "rec-005",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00056673,
      "latencyMs": 9936,
      "toolCalls": 4,
      "inputTokens": 24,
      "outputTokens": 454,
      "reasoningTokens": 249,
      "cacheReadTokens": 4280,
      "cacheWriteTokens": 820
    },
    {
      "trialId": "scored:rec-005:openai/gpt-5.6-luna:0",
      "fixtureId": "rec-005",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0009328000000000001,
      "latencyMs": 6558,
      "toolCalls": 3,
      "inputTokens": 3068,
      "outputTokens": 200,
      "reasoningTokens": 66,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-005:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "rec-005",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00014675400000000003,
      "latencyMs": 6413,
      "toolCalls": 3,
      "inputTokens": 764,
      "outputTokens": 322,
      "reasoningTokens": 0,
      "cacheReadTokens": 2816,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-005:anthropic/claude-opus-5:0",
      "fixtureId": "rec-005",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.02066525,
      "latencyMs": 8736,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 408,
      "reasoningTokens": 32,
      "cacheReadTokens": 4233,
      "cacheWriteTokens": 1203
    },
    {
      "trialId": "scored:rec-005:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "rec-005",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "duplicate_submission",
      "passed": false,
      "reasons": [
        "duplicate_submission"
      ],
      "cost": 0.004162708,
      "latencyMs": 9642,
      "toolCalls": 4,
      "inputTokens": 1501,
      "outputTokens": 468,
      "reasoningTokens": 44,
      "cacheReadTokens": 3497,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-005:openai/gpt-5.6-sol:0",
      "fixtureId": "rec-005",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01033,
      "latencyMs": 9006,
      "toolCalls": 4,
      "inputTokens": 3955,
      "outputTokens": 206,
      "reasoningTokens": 36,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-005:google/gemini-3.8-flash:0",
      "fixtureId": "rec-005",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00709125,
      "latencyMs": 12233,
      "toolCalls": 5,
      "inputTokens": 4855,
      "outputTokens": 274,
      "reasoningTokens": 646,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-005:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "rec-005",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.007179039999999999,
      "latencyMs": 20735,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 406,
      "reasoningTokens": 261,
      "cacheReadTokens": 4362,
      "cacheWriteTokens": 955
    },
    {
      "trialId": "scored:ext-006:openai/gpt-5.6-luna:0",
      "fixtureId": "ext-006",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00100522,
      "latencyMs": 8195,
      "toolCalls": 4,
      "inputTokens": 751,
      "outputTokens": 346,
      "reasoningTokens": 68,
      "cacheReadTokens": 2286,
      "cacheWriteTokens": 1250
    },
    {
      "trialId": "scored:ext-006:google/gemini-3.8-flash:0",
      "fixtureId": "ext-006",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.009520500000000001,
      "latencyMs": 12807,
      "toolCalls": 6,
      "inputTokens": 5699,
      "outputTokens": 476,
      "reasoningTokens": 923,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-006:anthropic/claude-opus-5:0",
      "fixtureId": "ext-006",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.03262825,
      "latencyMs": 12494,
      "toolCalls": 5,
      "inputTokens": 8,
      "outputTokens": 655,
      "reasoningTokens": 28,
      "cacheReadTokens": 7489,
      "cacheWriteTokens": 1883
    },
    {
      "trialId": "scored:ext-006:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "ext-006",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000263223,
      "latencyMs": 5809,
      "toolCalls": 4,
      "inputTokens": 1763,
      "outputTokens": 587,
      "reasoningTokens": 0,
      "cacheReadTokens": 2816,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-009:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "sup-009",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000445144,
      "latencyMs": 8756,
      "toolCalls": 8,
      "inputTokens": 2351,
      "outputTokens": 1156,
      "reasoningTokens": 0,
      "cacheReadTokens": 6400,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-006:openai/gpt-5.6-sol:0",
      "fixtureId": "ext-006",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0085474,
      "latencyMs": 11107,
      "toolCalls": 4,
      "inputTokens": 751,
      "outputTokens": 305,
      "reasoningTokens": 43,
      "cacheReadTokens": 2277,
      "cacheWriteTokens": 1244
    },
    {
      "trialId": "scored:ext-006:anthropic/claude-sonnet-5:0",
      "fixtureId": "ext-006",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.014587600000000001,
      "latencyMs": 14533,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 566,
      "reasoningTokens": 106,
      "cacheReadTokens": 4303,
      "cacheWriteTokens": 2798
    },
    {
      "trialId": "scored:ext-006:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "ext-006",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.004712091999999999,
      "latencyMs": 14547,
      "toolCalls": 4,
      "inputTokens": 1828,
      "outputTokens": 449,
      "reasoningTokens": 102,
      "cacheReadTokens": 2663,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-006:alibaba/qwen3.8-flash:0",
      "fixtureId": "ext-006",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0007787079999999999,
      "latencyMs": 15673,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 677,
      "reasoningTokens": 221,
      "cacheReadTokens": 5213,
      "cacheWriteTokens": 1347
    },
    {
      "trialId": "scored:sup-009:openai/gpt-5.6-luna:0",
      "fixtureId": "sup-009",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00129494,
      "latencyMs": 12118,
      "toolCalls": 6,
      "inputTokens": 892,
      "outputTokens": 388,
      "reasoningTokens": 141,
      "cacheReadTokens": 3062,
      "cacheWriteTokens": 1682
    },
    {
      "trialId": "scored:sup-009:alibaba/qwen3.8-flash:0",
      "fixtureId": "sup-009",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0014898020000000001,
      "latencyMs": 23213,
      "toolCalls": 9,
      "inputTokens": 30,
      "outputTokens": 1157,
      "reasoningTokens": 586,
      "cacheReadTokens": 8312,
      "cacheWriteTokens": 2664
    },
    {
      "trialId": "scored:sup-009:anthropic/claude-sonnet-5:0",
      "fixtureId": "sup-009",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0201111,
      "latencyMs": 12837,
      "toolCalls": 6,
      "inputTokens": 6,
      "outputTokens": 887,
      "reasoningTokens": 130,
      "cacheReadTokens": 5183,
      "cacheWriteTokens": 3557
    },
    {
      "trialId": "scored:sup-009:openai/gpt-5.6-sol:0",
      "fixtureId": "sup-009",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.011141,
      "latencyMs": 11316,
      "toolCalls": 6,
      "inputTokens": 892,
      "outputTokens": 395,
      "reasoningTokens": 66,
      "cacheReadTokens": 3035,
      "cacheWriteTokens": 1656
    },
    {
      "trialId": "scored:ext-006:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "ext-006",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01030133,
      "latencyMs": 29023,
      "toolCalls": 7,
      "inputTokens": 24,
      "outputTokens": 733,
      "reasoningTokens": 204,
      "cacheReadTokens": 5449,
      "cacheWriteTokens": 1482
    },
    {
      "trialId": "scored:sup-009:google/gemini-3.8-flash:0",
      "fixtureId": "sup-009",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.014073,
      "latencyMs": 19130,
      "toolCalls": 7,
      "inputTokens": 8624,
      "outputTokens": 326,
      "reasoningTokens": 1702,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-009:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "sup-009",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.007085495999999999,
      "latencyMs": 13751,
      "toolCalls": 6,
      "inputTokens": 2512,
      "outputTokens": 564,
      "reasoningTokens": 352,
      "cacheReadTokens": 3234,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-004:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "ext-004",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.006172539999999999,
      "latencyMs": 13815,
      "toolCalls": 7,
      "inputTokens": 2052,
      "outputTokens": 648,
      "reasoningTokens": 182,
      "cacheReadTokens": 4025,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-009:anthropic/claude-opus-5:0",
      "fixtureId": "sup-009",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.04832475,
      "latencyMs": 17744,
      "toolCalls": 6,
      "inputTokens": 6,
      "outputTokens": 1023,
      "reasoningTokens": 102,
      "cacheReadTokens": 5902,
      "cacheWriteTokens": 2755
    },
    {
      "trialId": "scored:ext-004:openai/gpt-5.6-sol:0",
      "fixtureId": "ext-004",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0078039,
      "latencyMs": 8142,
      "toolCalls": 6,
      "inputTokens": 742,
      "outputTokens": 265,
      "reasoningTokens": 37,
      "cacheReadTokens": 1037,
      "cacheWriteTokens": 1237
    },
    {
      "trialId": "scored:ext-004:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "ext-004",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000350146,
      "latencyMs": 8895,
      "toolCalls": 7,
      "inputTokens": 1893,
      "outputTokens": 950,
      "reasoningTokens": 0,
      "cacheReadTokens": 4352,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-004:openai/gpt-5.6-luna:0",
      "fixtureId": "ext-004",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00129637,
      "latencyMs": 7618,
      "toolCalls": 6,
      "inputTokens": 745,
      "outputTokens": 369,
      "reasoningTokens": 83,
      "cacheReadTokens": 1261,
      "cacheWriteTokens": 2319
    },
    {
      "trialId": "scored:ext-004:anthropic/claude-sonnet-5:0",
      "fixtureId": "ext-004",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0174974,
      "latencyMs": 11793,
      "toolCalls": 6,
      "inputTokens": 6,
      "outputTokens": 808,
      "reasoningTokens": 117,
      "cacheReadTokens": 4377,
      "cacheWriteTokens": 2944
    },
    {
      "trialId": "scored:ext-004:anthropic/claude-opus-5:0",
      "fixtureId": "ext-004",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.037414,
      "latencyMs": 12497,
      "toolCalls": 6,
      "inputTokens": 6,
      "outputTokens": 815,
      "reasoningTokens": 80,
      "cacheReadTokens": 5093,
      "cacheWriteTokens": 1994
    },
    {
      "trialId": "scored:ext-004:alibaba/qwen3.8-flash:0",
      "fixtureId": "ext-004",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0009619139999999999,
      "latencyMs": 15113,
      "toolCalls": 6,
      "inputTokens": 18,
      "outputTokens": 893,
      "reasoningTokens": 322,
      "cacheReadTokens": 3424,
      "cacheWriteTokens": 1666
    },
    {
      "trialId": "scored:ext-004:google/gemini-3.8-flash:0",
      "fixtureId": "ext-004",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0114855,
      "latencyMs": 20698,
      "toolCalls": 6,
      "inputTokens": 5019,
      "outputTokens": 343,
      "reasoningTokens": 1716,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-003:anthropic/claude-sonnet-5:0",
      "fixtureId": "sup-003",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": false,
      "reasons": [
        "params.decision: substantive value or membership differs"
      ],
      "cost": 0.0197698,
      "latencyMs": 17850,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 702,
      "reasoningTokens": 364,
      "cacheReadTokens": 4589,
      "cacheWriteTokens": 3272
    },
    {
      "trialId": "scored:sup-009:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "sup-009",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.020627629999999997,
      "latencyMs": 53401,
      "toolCalls": 9,
      "inputTokens": 30,
      "outputTokens": 1494,
      "reasoningTokens": 522,
      "cacheReadTokens": 8289,
      "cacheWriteTokens": 2825
    },
    {
      "trialId": "scored:sup-003:openai/gpt-5.6-sol:0",
      "fixtureId": "sup-003",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0090025,
      "latencyMs": 8389,
      "toolCalls": 4,
      "inputTokens": 846,
      "outputTokens": 278,
      "reasoningTokens": 68,
      "cacheReadTokens": 2465,
      "cacheWriteTokens": 1343
    },
    {
      "trialId": "scored:sup-003:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "sup-003",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00028299399999999995,
      "latencyMs": 5330,
      "toolCalls": 4,
      "inputTokens": 1649,
      "outputTokens": 726,
      "reasoningTokens": 0,
      "cacheReadTokens": 3328,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-003:openai/gpt-5.6-luna:0",
      "fixtureId": "sup-003",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00107682,
      "latencyMs": 9543,
      "toolCalls": 4,
      "inputTokens": 846,
      "outputTokens": 283,
      "reasoningTokens": 145,
      "cacheReadTokens": 2501,
      "cacheWriteTokens": 1376
    },
    {
      "trialId": "scored:sup-003:anthropic/claude-opus-5:0",
      "fixtureId": "sup-003",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.032896,
      "latencyMs": 11003,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 621,
      "reasoningTokens": 68,
      "cacheReadTokens": 5307,
      "cacheWriteTokens": 2078
    },
    {
      "trialId": "scored:ext-004:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "ext-004",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.012900430000000001,
      "latencyMs": 33556,
      "toolCalls": 8,
      "inputTokens": 24,
      "outputTokens": 883,
      "reasoningTokens": 402,
      "cacheReadTokens": 5279,
      "cacheWriteTokens": 1698
    },
    {
      "trialId": "scored:sup-003:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "sup-003",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.006065972,
      "latencyMs": 21959,
      "toolCalls": 5,
      "inputTokens": 2191,
      "outputTokens": 446,
      "reasoningTokens": 308,
      "cacheReadTokens": 4273,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-003:google/gemini-3.8-flash:0",
      "fixtureId": "sup-003",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01126725,
      "latencyMs": 13191,
      "toolCalls": 5,
      "inputTokens": 6858,
      "outputTokens": 337,
      "reasoningTokens": 1296,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-003:alibaba/qwen3.8-flash:0",
      "fixtureId": "sup-003",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0006273380000000001,
      "latencyMs": 9890,
      "toolCalls": 4,
      "inputTokens": 18,
      "outputTokens": 474,
      "reasoningTokens": 169,
      "cacheReadTokens": 3603,
      "cacheWriteTokens": 1323
    },
    {
      "trialId": "scored:rec-008:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "rec-008",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.004872428,
      "latencyMs": 10036,
      "toolCalls": 5,
      "inputTokens": 1822,
      "outputTokens": 435,
      "reasoningTokens": 146,
      "cacheReadTokens": 3787,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-008:openai/gpt-5.6-luna:0",
      "fixtureId": "rec-008",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00105778,
      "latencyMs": 9304,
      "toolCalls": 4,
      "inputTokens": 1664,
      "outputTokens": 286,
      "reasoningTokens": 66,
      "cacheReadTokens": 1104,
      "cacheWriteTokens": 1122
    },
    {
      "trialId": "scored:rec-008:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "rec-008",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00022467899999999998,
      "latencyMs": 10774,
      "toolCalls": 5,
      "inputTokens": 905,
      "outputTokens": 691,
      "reasoningTokens": 0,
      "cacheReadTokens": 3584,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-008:anthropic/claude-sonnet-5:0",
      "fixtureId": "rec-008",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0153591,
      "latencyMs": 14134,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 702,
      "reasoningTokens": 91,
      "cacheReadTokens": 4098,
      "cacheWriteTokens": 2639
    },
    {
      "trialId": "scored:sup-003:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "sup-003",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.013466120000000002,
      "latencyMs": 39001,
      "toolCalls": 6,
      "inputTokens": 30,
      "outputTokens": 756,
      "reasoningTokens": 483,
      "cacheReadTokens": 7336,
      "cacheWriteTokens": 1890
    },
    {
      "trialId": "scored:rec-008:openai/gpt-5.6-sol:0",
      "fixtureId": "rec-008",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.009819199999999998,
      "latencyMs": 9265,
      "toolCalls": 4,
      "inputTokens": 1664,
      "outputTokens": 307,
      "reasoningTokens": 39,
      "cacheReadTokens": 1106,
      "cacheWriteTokens": 1124
    },
    {
      "trialId": "scored:rec-008:anthropic/claude-opus-5:0",
      "fixtureId": "rec-008",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.03528775,
      "latencyMs": 12198,
      "toolCalls": 7,
      "inputTokens": 8,
      "outputTokens": 703,
      "reasoningTokens": 50,
      "cacheReadTokens": 7533,
      "cacheWriteTokens": 2025
    },
    {
      "trialId": "scored:rec-008:google/gemini-3.8-flash:0",
      "fixtureId": "rec-008",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.009079499999999999,
      "latencyMs": 16359,
      "toolCalls": 5,
      "inputTokens": 6441,
      "outputTokens": 260,
      "reasoningTokens": 873,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-008:alibaba/qwen3.8-flash:0",
      "fixtureId": "rec-008",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0007667500000000001,
      "latencyMs": 12577,
      "toolCalls": 7,
      "inputTokens": 30,
      "outputTokens": 646,
      "reasoningTokens": 203,
      "cacheReadTokens": 6320,
      "cacheWriteTokens": 1309
    },
    {
      "trialId": "scored:rec-007:anthropic/claude-sonnet-5:0",
      "fixtureId": "rec-007",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.014213100000000001,
      "latencyMs": 9207,
      "toolCalls": 5,
      "inputTokens": 6,
      "outputTokens": 602,
      "reasoningTokens": 55,
      "cacheReadTokens": 4218,
      "cacheWriteTokens": 2715
    },
    {
      "trialId": "scored:rec-007:alibaba/qwen3.8-flash:0",
      "fixtureId": "rec-007",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0007230759999999999,
      "latencyMs": 12612,
      "toolCalls": 6,
      "inputTokens": 24,
      "outputTokens": 620,
      "reasoningTokens": 186,
      "cacheReadTokens": 5051,
      "cacheWriteTokens": 1298
    },
    {
      "trialId": "scored:rec-008:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "rec-008",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01179307,
      "latencyMs": 31115,
      "toolCalls": 7,
      "inputTokens": 24,
      "outputTokens": 900,
      "reasoningTokens": 302,
      "cacheReadTokens": 5121,
      "cacheWriteTokens": 1465
    },
    {
      "trialId": "scored:rec-007:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "rec-007",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.005209556,
      "latencyMs": 11870,
      "toolCalls": 6,
      "inputTokens": 1878,
      "outputTokens": 551,
      "reasoningTokens": 95,
      "cacheReadTokens": 3919,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-007:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "rec-007",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000262042,
      "latencyMs": 5760,
      "toolCalls": 6,
      "inputTokens": 1164,
      "outputTokens": 666,
      "reasoningTokens": 0,
      "cacheReadTokens": 5120,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-007:openai/gpt-5.6-sol:0",
      "fixtureId": "rec-007",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0113497,
      "latencyMs": 16907,
      "toolCalls": 6,
      "inputTokens": 1748,
      "outputTokens": 358,
      "reasoningTokens": 62,
      "cacheReadTokens": 2331,
      "cacheWriteTokens": 1275
    },
    {
      "trialId": "scored:rec-007:openai/gpt-5.6-luna:0",
      "fixtureId": "rec-007",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0011389500000000001,
      "latencyMs": 8892,
      "toolCalls": 5,
      "inputTokens": 1749,
      "outputTokens": 312,
      "reasoningTokens": 73,
      "cacheReadTokens": 1195,
      "cacheWriteTokens": 1213
    },
    {
      "trialId": "scored:rec-007:anthropic/claude-opus-5:0",
      "fixtureId": "rec-007",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.029353249999999997,
      "latencyMs": 10773,
      "toolCalls": 5,
      "inputTokens": 6,
      "outputTokens": 585,
      "reasoningTokens": 42,
      "cacheReadTokens": 4934,
      "cacheWriteTokens": 1789
    },
    {
      "trialId": "scored:rec-007:google/gemini-3.8-flash:0",
      "fixtureId": "rec-007",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.009852,
      "latencyMs": 13210,
      "toolCalls": 5,
      "inputTokens": 4506,
      "outputTokens": 266,
      "reasoningTokens": 1460,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-009:openai/gpt-5.6-sol:0",
      "fixtureId": "rec-009",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0097482,
      "latencyMs": 8245,
      "toolCalls": 4,
      "inputTokens": 885,
      "outputTokens": 346,
      "reasoningTokens": 75,
      "cacheReadTokens": 2391,
      "cacheWriteTokens": 1316
    },
    {
      "trialId": "scored:rec-009:anthropic/claude-opus-5:0",
      "fixtureId": "rec-009",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0357615,
      "latencyMs": 11605,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 666,
      "reasoningTokens": 134,
      "cacheReadTokens": 5413,
      "cacheWriteTokens": 2084
    },
    {
      "trialId": "scored:rec-009:openai/gpt-5.6-luna:0",
      "fixtureId": "rec-009",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0011442000000000002,
      "latencyMs": 11102,
      "toolCalls": 4,
      "inputTokens": 885,
      "outputTokens": 359,
      "reasoningTokens": 128,
      "cacheReadTokens": 2415,
      "cacheWriteTokens": 1338
    },
    {
      "trialId": "scored:rec-009:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "rec-009",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00039019,
      "latencyMs": 7521,
      "toolCalls": 5,
      "inputTokens": 1885,
      "outputTokens": 1122,
      "reasoningTokens": 0,
      "cacheReadTokens": 5376,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-007:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "rec-007",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01221205,
      "latencyMs": 39633,
      "toolCalls": 9,
      "inputTokens": 24,
      "outputTokens": 1075,
      "reasoningTokens": 270,
      "cacheReadTokens": 5715,
      "cacheWriteTokens": 1249
    },
    {
      "trialId": "scored:rec-009:anthropic/claude-sonnet-5:0",
      "fixtureId": "rec-009",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0234257,
      "latencyMs": 20325,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 798,
      "reasoningTokens": 585,
      "cacheReadTokens": 4706,
      "cacheWriteTokens": 3457
    },
    {
      "trialId": "scored:rec-009:alibaba/qwen3.8-flash:0",
      "fixtureId": "rec-009",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00105049,
      "latencyMs": 15567,
      "toolCalls": 4,
      "inputTokens": 18,
      "outputTokens": 770,
      "reasoningTokens": 549,
      "cacheReadTokens": 3580,
      "cacheWriteTokens": 1852
    },
    {
      "trialId": "scored:rec-009:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "rec-009",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008035192,
      "latencyMs": 15956,
      "toolCalls": 5,
      "inputTokens": 2342,
      "outputTokens": 688,
      "reasoningTokens": 513,
      "cacheReadTokens": 4268,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-006:openai/gpt-5.6-luna:0",
      "fixtureId": "rec-006",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0009973500000000001,
      "latencyMs": 7731,
      "toolCalls": 4,
      "inputTokens": 1557,
      "outputTokens": 254,
      "reasoningTokens": 81,
      "cacheReadTokens": 1035,
      "cacheWriteTokens": 1053
    },
    {
      "trialId": "scored:rec-009:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "rec-009",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01337086,
      "latencyMs": 37546,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 909,
      "reasoningTokens": 472,
      "cacheReadTokens": 5658,
      "cacheWriteTokens": 1630
    },
    {
      "trialId": "scored:rec-006:anthropic/claude-opus-5:0",
      "fixtureId": "rec-006",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0251435,
      "latencyMs": 9597,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 502,
      "reasoningTokens": 37,
      "cacheReadTokens": 4602,
      "cacheWriteTokens": 1494
    },
    {
      "trialId": "scored:rec-006:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "rec-006",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.004457464,
      "latencyMs": 9898,
      "toolCalls": 5,
      "inputTokens": 1676,
      "outputTokens": 421,
      "reasoningTokens": 106,
      "cacheReadTokens": 3596,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-006:anthropic/claude-sonnet-5:0",
      "fixtureId": "rec-006",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.013206699999999998,
      "latencyMs": 13231,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 553,
      "reasoningTokens": 77,
      "cacheReadTokens": 3886,
      "cacheWriteTokens": 2447
    },
    {
      "trialId": "scored:rec-006:alibaba/qwen3.8-flash:0",
      "fixtureId": "rec-006",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000667186,
      "latencyMs": 13388,
      "toolCalls": 5,
      "inputTokens": 18,
      "outputTokens": 609,
      "reasoningTokens": 174,
      "cacheReadTokens": 3206,
      "cacheWriteTokens": 1225
    },
    {
      "trialId": "scored:rec-006:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "rec-006",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00033526,
      "latencyMs": 8102,
      "toolCalls": 7,
      "inputTokens": 1539,
      "outputTokens": 888,
      "reasoningTokens": 0,
      "cacheReadTokens": 5888,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-006:openai/gpt-5.6-sol:0",
      "fixtureId": "rec-006",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.010668,
      "latencyMs": 8236,
      "toolCalls": 4,
      "inputTokens": 2579,
      "outputTokens": 254,
      "reasoningTokens": 37,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 1040
    },
    {
      "trialId": "scored:rec-004:anthropic/claude-sonnet-5:0",
      "fixtureId": "rec-004",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0134519,
      "latencyMs": 9940,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 571,
      "reasoningTokens": 9,
      "cacheReadTokens": 4337,
      "cacheWriteTokens": 2709
    },
    {
      "trialId": "scored:rec-004:anthropic/claude-opus-5:0",
      "fixtureId": "rec-004",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.03198925,
      "latencyMs": 10981,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 655,
      "reasoningTokens": 49,
      "cacheReadTokens": 5056,
      "cacheWriteTokens": 1893
    },
    {
      "trialId": "scored:rec-009:google/gemini-3.8-flash:0",
      "fixtureId": "rec-009",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.018349499999999998,
      "latencyMs": 25231,
      "toolCalls": 4,
      "inputTokens": 6426,
      "outputTokens": 462,
      "reasoningTokens": 3146,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-004:openai/gpt-5.6-luna:0",
      "fixtureId": "rec-004",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0009422100000000001,
      "latencyMs": 8081,
      "toolCalls": 4,
      "inputTokens": 782,
      "outputTokens": 318,
      "reasoningTokens": 41,
      "cacheReadTokens": 2263,
      "cacheWriteTokens": 1239
    },
    {
      "trialId": "scored:rec-004:alibaba/qwen3.8-flash:0",
      "fixtureId": "rec-004",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0006357819999999999,
      "latencyMs": 10183,
      "toolCalls": 4,
      "inputTokens": 18,
      "outputTokens": 345,
      "reasoningTokens": 304,
      "cacheReadTokens": 3467,
      "cacheWriteTokens": 1362
    },
    {
      "trialId": "scored:rec-004:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "rec-004",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000335785,
      "latencyMs": 7454,
      "toolCalls": 6,
      "inputTokens": 1805,
      "outputTokens": 853,
      "reasoningTokens": 0,
      "cacheReadTokens": 4864,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-004:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "rec-004",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.004732728,
      "latencyMs": 12266,
      "toolCalls": 5,
      "inputTokens": 1864,
      "outputTokens": 471,
      "reasoningTokens": 58,
      "cacheReadTokens": 4032,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-006:google/gemini-3.8-flash:0",
      "fixtureId": "rec-006",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.011769,
      "latencyMs": 27463,
      "toolCalls": 7,
      "inputTokens": 7717,
      "outputTokens": 387,
      "reasoningTokens": 1208,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-004:google/gemini-3.8-flash:0",
      "fixtureId": "rec-004",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0102165,
      "latencyMs": 13200,
      "toolCalls": 5,
      "inputTokens": 5537,
      "outputTokens": 469,
      "reasoningTokens": 1148,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-006:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "rec-006",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01049364,
      "latencyMs": 28743,
      "toolCalls": 6,
      "inputTokens": 24,
      "outputTokens": 809,
      "reasoningTokens": 232,
      "cacheReadTokens": 4792,
      "cacheWriteTokens": 1354
    },
    {
      "trialId": "scored:ext-002:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "ext-002",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000303858,
      "latencyMs": 4807,
      "toolCalls": 5,
      "inputTokens": 1705,
      "outputTokens": 694,
      "reasoningTokens": 0,
      "cacheReadTokens": 4864,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-004:openai/gpt-5.6-sol:0",
      "fixtureId": "rec-004",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0085971,
      "latencyMs": 7666,
      "toolCalls": 4,
      "inputTokens": 782,
      "outputTokens": 320,
      "reasoningTokens": 31,
      "cacheReadTokens": 2253,
      "cacheWriteTokens": 1229
    },
    {
      "trialId": "scored:ext-002:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "ext-002",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.004909916,
      "latencyMs": 12882,
      "toolCalls": 4,
      "inputTokens": 1895,
      "outputTokens": 458,
      "reasoningTokens": 120,
      "cacheReadTokens": 2719,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-002:alibaba/qwen3.8-flash:0",
      "fixtureId": "ext-002",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000765986,
      "latencyMs": 13717,
      "toolCalls": 4,
      "inputTokens": 24,
      "outputTokens": 544,
      "reasoningTokens": 307,
      "cacheReadTokens": 5236,
      "cacheWriteTokens": 1392
    },
    {
      "trialId": "scored:ext-002:openai/gpt-5.6-luna:0",
      "fixtureId": "ext-002",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00095438,
      "latencyMs": 7552,
      "toolCalls": 3,
      "inputTokens": 785,
      "outputTokens": 315,
      "reasoningTokens": 46,
      "cacheReadTokens": 2309,
      "cacheWriteTokens": 1272
    },
    {
      "trialId": "scored:ext-002:anthropic/claude-sonnet-5:0",
      "fixtureId": "ext-002",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0153627,
      "latencyMs": 13314,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 690,
      "reasoningTokens": 74,
      "cacheReadTokens": 4266,
      "cacheWriteTokens": 2743
    },
    {
      "trialId": "scored:rec-004:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "rec-004",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01261642,
      "latencyMs": 34033,
      "toolCalls": 7,
      "inputTokens": 30,
      "outputTokens": 857,
      "reasoningTokens": 368,
      "cacheReadTokens": 7376,
      "cacheWriteTokens": 1581
    },
    {
      "trialId": "scored:ext-002:openai/gpt-5.6-sol:0",
      "fixtureId": "ext-002",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0088817,
      "latencyMs": 10679,
      "toolCalls": 3,
      "inputTokens": 785,
      "outputTokens": 340,
      "reasoningTokens": 31,
      "cacheReadTokens": 2296,
      "cacheWriteTokens": 1257
    },
    {
      "trialId": "scored:ext-002:google/gemini-3.8-flash:0",
      "fixtureId": "ext-002",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0075727500000000005,
      "latencyMs": 10912,
      "toolCalls": 3,
      "inputTokens": 4057,
      "outputTokens": 246,
      "reasoningTokens": 962,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-002:anthropic/claude-opus-5:0",
      "fixtureId": "ext-002",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.02843925,
      "latencyMs": 11264,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 573,
      "reasoningTokens": 20,
      "cacheReadTokens": 4981,
      "cacheWriteTokens": 1775
    },
    {
      "trialId": "scored:ext-002:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "ext-002",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01423568,
      "latencyMs": 36976,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 994,
      "reasoningTokens": 439,
      "cacheReadTokens": 5204,
      "cacheWriteTokens": 1882
    },
    {
      "trialId": "scored:sup-005:openai/gpt-5.6-sol:0",
      "fixtureId": "sup-005",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.010974600000000001,
      "latencyMs": 13699,
      "toolCalls": 6,
      "inputTokens": 962,
      "outputTokens": 335,
      "reasoningTokens": 78,
      "cacheReadTokens": 4403,
      "cacheWriteTokens": 1616
    },
    {
      "trialId": "scored:sup-005:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "sup-005",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008273804000000001,
      "latencyMs": 15209,
      "toolCalls": 8,
      "inputTokens": 2725,
      "outputTokens": 638,
      "reasoningTokens": 465,
      "cacheReadTokens": 7021,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-005:alibaba/qwen3.8-flash:0",
      "fixtureId": "sup-005",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.001276072,
      "latencyMs": 18827,
      "toolCalls": 7,
      "inputTokens": 24,
      "outputTokens": 810,
      "reasoningTokens": 682,
      "cacheReadTokens": 6662,
      "cacheWriteTokens": 2322
    },
    {
      "trialId": "scored:sup-005:anthropic/claude-sonnet-5:0",
      "fixtureId": "sup-005",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0231835,
      "latencyMs": 21842,
      "toolCalls": 6,
      "inputTokens": 8,
      "outputTokens": 803,
      "reasoningTokens": 421,
      "cacheReadTokens": 8500,
      "cacheWriteTokens": 3691
    },
    {
      "trialId": "scored:sup-005:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "sup-005",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00041320699999999996,
      "latencyMs": 8713,
      "toolCalls": 6,
      "inputTokens": 1636,
      "outputTokens": 1279,
      "reasoningTokens": 0,
      "cacheReadTokens": 6656,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-005:openai/gpt-5.6-luna:0",
      "fixtureId": "sup-005",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00128799,
      "latencyMs": 10319,
      "toolCalls": 6,
      "inputTokens": 962,
      "outputTokens": 344,
      "reasoningTokens": 150,
      "cacheReadTokens": 4452,
      "cacheWriteTokens": 1655
    },
    {
      "trialId": "scored:ext-009:anthropic/claude-sonnet-5:0",
      "fixtureId": "ext-009",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0108405,
      "latencyMs": 9999,
      "toolCalls": 2,
      "inputTokens": 6,
      "outputTokens": 463,
      "reasoningTokens": 0,
      "cacheReadTokens": 3605,
      "cacheWriteTokens": 2191
    },
    {
      "trialId": "scored:ext-009:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "ext-009",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.003757864,
      "latencyMs": 7679,
      "toolCalls": 2,
      "inputTokens": 1517,
      "outputTokens": 334,
      "reasoningTokens": 84,
      "cacheReadTokens": 2276,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-009:anthropic/claude-opus-5:0",
      "fixtureId": "ext-009",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.021728499999999998,
      "latencyMs": 8390,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 437,
      "reasoningTokens": 0,
      "cacheReadTokens": 4422,
      "cacheWriteTokens": 1370
    },
    {
      "trialId": "scored:sup-005:google/gemini-3.8-flash:0",
      "fixtureId": "sup-005",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01933275,
      "latencyMs": 19789,
      "toolCalls": 8,
      "inputTokens": 14082,
      "outputTokens": 493,
      "reasoningTokens": 1846,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-005:anthropic/claude-opus-5:0",
      "fixtureId": "sup-005",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.04611825,
      "latencyMs": 15465,
      "toolCalls": 6,
      "inputTokens": 8,
      "outputTokens": 830,
      "reasoningTokens": 184,
      "cacheReadTokens": 8994,
      "cacheWriteTokens": 2597
    },
    {
      "trialId": "scored:ext-009:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "ext-009",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00018776100000000002,
      "latencyMs": 4115,
      "toolCalls": 3,
      "inputTokens": 888,
      "outputTokens": 505,
      "reasoningTokens": 0,
      "cacheReadTokens": 3072,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-009:openai/gpt-5.6-sol:0",
      "fixtureId": "ext-009",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.006402,
      "latencyMs": 6307,
      "toolCalls": 2,
      "inputTokens": 2391,
      "outputTokens": 149,
      "reasoningTokens": 13,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-009:alibaba/qwen3.8-flash:0",
      "fixtureId": "ext-009",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000462154,
      "latencyMs": 7077,
      "toolCalls": 2,
      "inputTokens": 18,
      "outputTokens": 388,
      "reasoningTokens": 115,
      "cacheReadTokens": 3054,
      "cacheWriteTokens": 870
    },
    {
      "trialId": "scored:ext-009:google/gemini-3.8-flash:0",
      "fixtureId": "ext-009",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00576075,
      "latencyMs": 7577,
      "toolCalls": 2,
      "inputTokens": 3106,
      "outputTokens": 185,
      "reasoningTokens": 730,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-009:openai/gpt-5.6-luna:0",
      "fixtureId": "ext-009",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0010362,
      "latencyMs": 8461,
      "toolCalls": 2,
      "inputTokens": 3381,
      "outputTokens": 237,
      "reasoningTokens": 63,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-005:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "sup-005",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01552355,
      "latencyMs": 36862,
      "toolCalls": 9,
      "inputTokens": 24,
      "outputTokens": 895,
      "reasoningTokens": 573,
      "cacheReadTokens": 6765,
      "cacheWriteTokens": 2207
    },
    {
      "trialId": "scored:sup-007:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "sup-007",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00036387,
      "latencyMs": 8560,
      "toolCalls": 8,
      "inputTokens": 1870,
      "outputTokens": 934,
      "reasoningTokens": 0,
      "cacheReadTokens": 5632,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-009:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "ext-009",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0067686199999999995,
      "latencyMs": 19206,
      "toolCalls": 3,
      "inputTokens": 18,
      "outputTokens": 510,
      "reasoningTokens": 128,
      "cacheReadTokens": 3086,
      "cacheWriteTokens": 952
    },
    {
      "trialId": "scored:sup-007:openai/gpt-5.6-sol:0",
      "fixtureId": "sup-007",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0091693,
      "latencyMs": 11473,
      "toolCalls": 6,
      "inputTokens": 795,
      "outputTokens": 290,
      "reasoningTokens": 56,
      "cacheReadTokens": 2684,
      "cacheWriteTokens": 1433
    },
    {
      "trialId": "scored:sup-007:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "sup-007",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.006188424,
      "latencyMs": 12271,
      "toolCalls": 7,
      "inputTokens": 2244,
      "outputTokens": 564,
      "reasoningTokens": 202,
      "cacheReadTokens": 4386,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-007:google/gemini-3.8-flash:0",
      "fixtureId": "sup-007",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.014081999999999999,
      "latencyMs": 16011,
      "toolCalls": 8,
      "inputTokens": 9986,
      "outputTokens": 360,
      "reasoningTokens": 1398,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-007:anthropic/claude-opus-5:0",
      "fixtureId": "sup-007",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.04889575,
      "latencyMs": 16547,
      "toolCalls": 8,
      "inputTokens": 8,
      "outputTokens": 924,
      "reasoningTokens": 189,
      "cacheReadTokens": 8549,
      "cacheWriteTokens": 2681
    },
    {
      "trialId": "scored:sup-007:openai/gpt-5.6-luna:0",
      "fixtureId": "sup-007",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0011802599999999998,
      "latencyMs": 9200,
      "toolCalls": 6,
      "inputTokens": 795,
      "outputTokens": 282,
      "reasoningTokens": 200,
      "cacheReadTokens": 2843,
      "cacheWriteTokens": 1544
    },
    {
      "trialId": "scored:sup-007:alibaba/qwen3.8-flash:0",
      "fixtureId": "sup-007",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00131515,
      "latencyMs": 25957,
      "toolCalls": 9,
      "inputTokens": 60,
      "outputTokens": 792,
      "reasoningTokens": 521,
      "cacheReadTokens": 17165,
      "cacheWriteTokens": 2069
    },
    {
      "trialId": "scored:sup-007:anthropic/claude-sonnet-5:0",
      "fixtureId": "sup-007",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.017259,
      "latencyMs": 13351,
      "toolCalls": 6,
      "inputTokens": 6,
      "outputTokens": 748,
      "reasoningTokens": 93,
      "cacheReadTokens": 4785,
      "cacheWriteTokens": 3152
    },
    {
      "trialId": "scored:sup-008:openai/gpt-5.6-luna:0",
      "fixtureId": "sup-008",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0017413499999999998,
      "latencyMs": 10689,
      "toolCalls": 7,
      "inputTokens": 12,
      "outputTokens": 515,
      "reasoningTokens": 129,
      "cacheReadTokens": 3020,
      "cacheWriteTokens": 3623
    },
    {
      "trialId": "scored:sup-008:anthropic/claude-sonnet-5:0",
      "fixtureId": "sup-008",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0285387,
      "latencyMs": 18814,
      "toolCalls": 7,
      "inputTokens": 6,
      "outputTokens": 1259,
      "reasoningTokens": 348,
      "cacheReadTokens": 5896,
      "cacheWriteTokens": 4511
    },
    {
      "trialId": "scored:sup-008:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "sup-008",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000504842,
      "latencyMs": 8267,
      "toolCalls": 8,
      "inputTokens": 3489,
      "outputTokens": 1262,
      "reasoningTokens": 0,
      "cacheReadTokens": 3328,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-008:anthropic/claude-opus-5:0",
      "fixtureId": "sup-008",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.056775,
      "latencyMs": 18539,
      "toolCalls": 7,
      "inputTokens": 6,
      "outputTokens": 1177,
      "reasoningTokens": 115,
      "cacheReadTokens": 6615,
      "cacheWriteTokens": 3382
    },
    {
      "trialId": "scored:sup-008:google/gemini-3.8-flash:0",
      "fixtureId": "sup-008",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01726425,
      "latencyMs": 21641,
      "toolCalls": 7,
      "inputTokens": 8504,
      "outputTokens": 410,
      "reasoningTokens": 2493,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-008:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "sup-008",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.009095064,
      "latencyMs": 17165,
      "toolCalls": 8,
      "inputTokens": 3046,
      "outputTokens": 756,
      "reasoningTokens": 466,
      "cacheReadTokens": 5346,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-007:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "sup-007",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01462546,
      "latencyMs": 39895,
      "toolCalls": 10,
      "inputTokens": 36,
      "outputTokens": 869,
      "reasoningTokens": 434,
      "cacheReadTokens": 9988,
      "cacheWriteTokens": 2015
    },
    {
      "trialId": "scored:ext-005:openai/gpt-5.6-sol:0",
      "fixtureId": "ext-005",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.010617499999999998,
      "latencyMs": 9563,
      "toolCalls": 4,
      "inputTokens": 2575,
      "outputTokens": 251,
      "reasoningTokens": 37,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 1035
    },
    {
      "trialId": "scored:sup-008:openai/gpt-5.6-sol:0",
      "fixtureId": "sup-008",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.011084899999999998,
      "latencyMs": 11433,
      "toolCalls": 7,
      "inputTokens": 12,
      "outputTokens": 471,
      "reasoningTokens": 66,
      "cacheReadTokens": 4567,
      "cacheWriteTokens": 1911
    },
    {
      "trialId": "scored:sup-008:alibaba/qwen3.8-flash:0",
      "fixtureId": "sup-008",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.001825456,
      "latencyMs": 26177,
      "toolCalls": 8,
      "inputTokens": 24,
      "outputTokens": 1490,
      "reasoningTokens": 730,
      "cacheReadTokens": 7351,
      "cacheWriteTokens": 3303
    },
    {
      "trialId": "scored:ext-005:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "ext-005",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000251331,
      "latencyMs": 4923,
      "toolCalls": 5,
      "inputTokens": 1388,
      "outputTokens": 555,
      "reasoningTokens": 0,
      "cacheReadTokens": 4352,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-005:anthropic/claude-sonnet-5:0",
      "fixtureId": "ext-005",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.013347200000000002,
      "latencyMs": 9781,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 562,
      "reasoningTokens": 81,
      "cacheReadTokens": 3876,
      "cacheWriteTokens": 2452
    },
    {
      "trialId": "scored:ext-005:google/gemini-3.8-flash:0",
      "fixtureId": "ext-005",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01078875,
      "latencyMs": 13474,
      "toolCalls": 5,
      "inputTokens": 6170,
      "outputTokens": 197,
      "reasoningTokens": 1446,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-005:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "ext-005",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.004972396,
      "latencyMs": 13657,
      "toolCalls": 5,
      "inputTokens": 1743,
      "outputTokens": 445,
      "reasoningTokens": 186,
      "cacheReadTokens": 3929,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-005:openai/gpt-5.6-luna:0",
      "fixtureId": "ext-005",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0009963300000000001,
      "latencyMs": 7275,
      "toolCalls": 4,
      "inputTokens": 1560,
      "outputTokens": 259,
      "reasoningTokens": 76,
      "cacheReadTokens": 1029,
      "cacheWriteTokens": 1047
    },
    {
      "trialId": "scored:ext-005:anthropic/claude-opus-5:0",
      "fixtureId": "ext-005",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.030905249999999995,
      "latencyMs": 14742,
      "toolCalls": 5,
      "inputTokens": 8,
      "outputTokens": 640,
      "reasoningTokens": 61,
      "cacheReadTokens": 6818,
      "cacheWriteTokens": 1589
    },
    {
      "trialId": "scored:sup-006:anthropic/claude-sonnet-5:0",
      "fixtureId": "sup-006",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0156407,
      "latencyMs": 10682,
      "toolCalls": 5,
      "inputTokens": 6,
      "outputTokens": 661,
      "reasoningTokens": 59,
      "cacheReadTokens": 4706,
      "cacheWriteTokens": 2995
    },
    {
      "trialId": "scored:sup-008:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "sup-008",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.02395124,
      "latencyMs": 58839,
      "toolCalls": 9,
      "inputTokens": 30,
      "outputTokens": 1489,
      "reasoningTokens": 804,
      "cacheReadTokens": 9372,
      "cacheWriteTokens": 3416
    },
    {
      "trialId": "scored:sup-006:anthropic/claude-opus-5:0",
      "fixtureId": "sup-006",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.037836999999999996,
      "latencyMs": 12842,
      "toolCalls": 6,
      "inputTokens": 8,
      "outputTokens": 743,
      "reasoningTokens": 63,
      "cacheReadTokens": 8169,
      "cacheWriteTokens": 2170
    },
    {
      "trialId": "scored:sup-006:openai/gpt-5.6-luna:0",
      "fixtureId": "sup-006",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0010076899999999999,
      "latencyMs": 7384,
      "toolCalls": 5,
      "inputTokens": 836,
      "outputTokens": 270,
      "reasoningTokens": 99,
      "cacheReadTokens": 2572,
      "cacheWriteTokens": 1385
    },
    {
      "trialId": "scored:ext-005:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "ext-005",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00980175,
      "latencyMs": 28707,
      "toolCalls": 6,
      "inputTokens": 24,
      "outputTokens": 592,
      "reasoningTokens": 392,
      "cacheReadTokens": 4925,
      "cacheWriteTokens": 1205
    },
    {
      "trialId": "scored:sup-006:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "sup-006",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000383152,
      "latencyMs": 7459,
      "toolCalls": 7,
      "inputTokens": 1862,
      "outputTokens": 1064,
      "reasoningTokens": 0,
      "cacheReadTokens": 5632,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-006:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "sup-006",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.005320304,
      "latencyMs": 12039,
      "toolCalls": 6,
      "inputTokens": 2071,
      "outputTokens": 478,
      "reasoningTokens": 127,
      "cacheReadTokens": 4336,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-005:alibaba/qwen3.8-flash:0",
      "fixtureId": "ext-005",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0007142040000000001,
      "latencyMs": 38597,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 604,
      "reasoningTokens": 250,
      "cacheReadTokens": 4749,
      "cacheWriteTokens": 1165
    },
    {
      "trialId": "scored:sup-006:google/gemini-3.8-flash:0",
      "fixtureId": "sup-006",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.009129,
      "latencyMs": 12823,
      "toolCalls": 6,
      "inputTokens": 6072,
      "outputTokens": 240,
      "reasoningTokens": 980,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-006:openai/gpt-5.6-sol:0",
      "fixtureId": "sup-006",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0086984,
      "latencyMs": 7522,
      "toolCalls": 5,
      "inputTokens": 836,
      "outputTokens": 287,
      "reasoningTokens": 28,
      "cacheReadTokens": 2532,
      "cacheWriteTokens": 1348
    },
    {
      "trialId": "scored:sup-006:alibaba/qwen3.8-flash:0",
      "fixtureId": "sup-006",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0009280219999999999,
      "latencyMs": 14439,
      "toolCalls": 6,
      "inputTokens": 24,
      "outputTokens": 702,
      "reasoningTokens": 335,
      "cacheReadTokens": 5787,
      "cacheWriteTokens": 1721
    },
    {
      "trialId": "scored:ext-008:anthropic/claude-opus-5:0",
      "fixtureId": "ext-008",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01918425,
      "latencyMs": 7880,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 322,
      "reasoningTokens": 39,
      "cacheReadTokens": 4346,
      "cacheWriteTokens": 1273
    },
    {
      "trialId": "scored:ext-008:deepseek/deepseek-v4-pro-0813:0",
      "fixtureId": "ext-008",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.003934567999999999,
      "latencyMs": 9514,
      "toolCalls": 3,
      "inputTokens": 1477,
      "outputTokens": 376,
      "reasoningTokens": 88,
      "cacheReadTokens": 3352,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-006:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "sup-006",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.011919160000000002,
      "latencyMs": 33893,
      "toolCalls": 7,
      "inputTokens": 24,
      "outputTokens": 755,
      "reasoningTokens": 363,
      "cacheReadTokens": 5798,
      "cacheWriteTokens": 1671
    },
    {
      "trialId": "scored:ext-008:openai/gpt-5.6-luna:0",
      "fixtureId": "ext-008",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0009362000000000001,
      "latencyMs": 6209,
      "toolCalls": 2,
      "inputTokens": 3211,
      "outputTokens": 174,
      "reasoningTokens": 71,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-008:openai/gpt-5.6-sol:0",
      "fixtureId": "ext-008",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008612,
      "latencyMs": 7617,
      "toolCalls": 2,
      "inputTokens": 3201,
      "outputTokens": 186,
      "reasoningTokens": 35,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-008:anthropic/claude-sonnet-5:0",
      "fixtureId": "ext-008",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0114859,
      "latencyMs": 12281,
      "toolCalls": 2,
      "inputTokens": 6,
      "outputTokens": 436,
      "reasoningTokens": 101,
      "cacheReadTokens": 3532,
      "cacheWriteTokens": 2159
    },
    {
      "trialId": "scored:ext-008:deepseek/deepseek-v4-flash-0731:0",
      "fixtureId": "ext-008",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000283258,
      "latencyMs": 7825,
      "toolCalls": 4,
      "inputTokens": 1687,
      "outputTokens": 662,
      "reasoningTokens": 0,
      "cacheReadTokens": 3840,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-003:openai/gpt-5.6-luna:1",
      "fixtureId": "sup-003",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00104171,
      "latencyMs": 7833,
      "toolCalls": 4,
      "inputTokens": 846,
      "outputTokens": 279,
      "reasoningTokens": 120,
      "cacheReadTokens": 2498,
      "cacheWriteTokens": 1375
    },
    {
      "trialId": "scored:ext-008:google/gemini-3.8-flash:0",
      "fixtureId": "ext-008",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.007617,
      "latencyMs": 10362,
      "toolCalls": 3,
      "inputTokens": 4821,
      "outputTokens": 145,
      "reasoningTokens": 922,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-003:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "sup-003",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000280167,
      "latencyMs": 7337,
      "toolCalls": 5,
      "inputTokens": 1152,
      "outputTokens": 767,
      "reasoningTokens": 0,
      "cacheReadTokens": 5376,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-008:alibaba/qwen3.8-flash:0",
      "fixtureId": "ext-008",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0005792600000000001,
      "latencyMs": 19761,
      "toolCalls": 3,
      "inputTokens": 24,
      "outputTokens": 467,
      "reasoningTokens": 207,
      "cacheReadTokens": 4465,
      "cacheWriteTokens": 936
    },
    {
      "trialId": "scored:sup-003:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "sup-003",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.007712056,
      "latencyMs": 11805,
      "toolCalls": 5,
      "inputTokens": 3462,
      "outputTokens": 440,
      "reasoningTokens": 335,
      "cacheReadTokens": 1664,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-003:anthropic/claude-opus-5:1",
      "fixtureId": "sup-003",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.041159,
      "latencyMs": 12887,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 904,
      "reasoningTokens": 60,
      "cacheReadTokens": 5308,
      "cacheWriteTokens": 2300
    },
    {
      "trialId": "scored:ext-008:alibaba/qwen3.8-max-0902:0",
      "fixtureId": "ext-008",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00804201,
      "latencyMs": 21753,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 581,
      "reasoningTokens": 204,
      "cacheReadTokens": 4553,
      "cacheWriteTokens": 1004
    },
    {
      "trialId": "scored:sup-003:anthropic/claude-sonnet-5:1",
      "fixtureId": "sup-003",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0182348,
      "latencyMs": 14143,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 578,
      "reasoningTokens": 344,
      "cacheReadTokens": 4589,
      "cacheWriteTokens": 3234
    },
    {
      "trialId": "scored:sup-003:openai/gpt-5.6-sol:1",
      "fixtureId": "sup-003",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0089873,
      "latencyMs": 11532,
      "toolCalls": 4,
      "inputTokens": 846,
      "outputTokens": 282,
      "reasoningTokens": 63,
      "cacheReadTokens": 2464,
      "cacheWriteTokens": 1341
    },
    {
      "trialId": "scored:sup-003:google/gemini-3.8-flash:1",
      "fixtureId": "sup-003",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.009528000000000002,
      "latencyMs": 12412,
      "toolCalls": 4,
      "inputTokens": 4719,
      "outputTokens": 203,
      "reasoningTokens": 1394,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-007:openai/gpt-5.6-sol:1",
      "fixtureId": "ext-007",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0083341,
      "latencyMs": 10207,
      "toolCalls": 4,
      "inputTokens": 837,
      "outputTokens": 284,
      "reasoningTokens": 33,
      "cacheReadTokens": 2238,
      "cacheWriteTokens": 1217
    },
    {
      "trialId": "scored:ext-007:alibaba/qwen3.8-flash:1",
      "fixtureId": "ext-007",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000722592,
      "latencyMs": 11310,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 531,
      "reasoningTokens": 265,
      "cacheReadTokens": 5152,
      "cacheWriteTokens": 1311
    },
    {
      "trialId": "scored:ext-007:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "ext-007",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000240994,
      "latencyMs": 5013,
      "toolCalls": 5,
      "inputTokens": 833,
      "outputTokens": 646,
      "reasoningTokens": 0,
      "cacheReadTokens": 5632,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-007:google/gemini-3.8-flash:1",
      "fixtureId": "ext-007",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00942375,
      "latencyMs": 15751,
      "toolCalls": 6,
      "inputTokens": 7045,
      "outputTokens": 265,
      "reasoningTokens": 839,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-003:alibaba/qwen3.8-flash:1",
      "fixtureId": "sup-003",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.001223764,
      "latencyMs": 22354,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 865,
      "reasoningTokens": 629,
      "cacheReadTokens": 5909,
      "cacheWriteTokens": 2116
    },
    {
      "trialId": "scored:ext-007:openai/gpt-5.6-luna:1",
      "fixtureId": "ext-007",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0009070399999999999,
      "latencyMs": 7156,
      "toolCalls": 4,
      "inputTokens": 837,
      "outputTokens": 279,
      "reasoningTokens": 44,
      "cacheReadTokens": 2252,
      "cacheWriteTokens": 1228
    },
    {
      "trialId": "scored:ext-007:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "ext-007",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.004690136000000001,
      "latencyMs": 10603,
      "toolCalls": 5,
      "inputTokens": 1868,
      "outputTokens": 418,
      "reasoningTokens": 98,
      "cacheReadTokens": 4114,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-002:alibaba/qwen3.8-flash:1",
      "fixtureId": "sup-002",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000655238,
      "latencyMs": 10804,
      "toolCalls": 4,
      "inputTokens": 18,
      "outputTokens": 535,
      "reasoningTokens": 166,
      "cacheReadTokens": 3543,
      "cacheWriteTokens": 1331
    },
    {
      "trialId": "scored:ext-007:anthropic/claude-opus-5:1",
      "fixtureId": "ext-007",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.03522325,
      "latencyMs": 14912,
      "toolCalls": 6,
      "inputTokens": 8,
      "outputTokens": 708,
      "reasoningTokens": 21,
      "cacheReadTokens": 7954,
      "cacheWriteTokens": 2077
    },
    {
      "trialId": "scored:ext-007:anthropic/claude-sonnet-5:1",
      "fixtureId": "ext-007",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0141936,
      "latencyMs": 12014,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 558,
      "reasoningTokens": 77,
      "cacheReadTokens": 4458,
      "cacheWriteTokens": 2776
    },
    {
      "trialId": "scored:sup-003:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "sup-003",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01630403,
      "latencyMs": 46961,
      "toolCalls": 6,
      "inputTokens": 30,
      "outputTokens": 878,
      "reasoningTokens": 708,
      "cacheReadTokens": 7459,
      "cacheWriteTokens": 2184
    },
    {
      "trialId": "scored:sup-002:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "sup-002",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00018847599999999998,
      "latencyMs": 4344,
      "toolCalls": 4,
      "inputTokens": 590,
      "outputTokens": 564,
      "reasoningTokens": 0,
      "cacheReadTokens": 4096,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-007:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "ext-007",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008458180000000001,
      "latencyMs": 27891,
      "toolCalls": 6,
      "inputTokens": 30,
      "outputTokens": 659,
      "reasoningTokens": 207,
      "cacheReadTokens": 7454,
      "cacheWriteTokens": 774
    },
    {
      "trialId": "scored:sup-002:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "sup-002",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.004105332000000001,
      "latencyMs": 7047,
      "toolCalls": 4,
      "inputTokens": 1789,
      "outputTokens": 343,
      "reasoningTokens": 67,
      "cacheReadTokens": 2733,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-002:anthropic/claude-opus-5:1",
      "fixtureId": "sup-002",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0344675,
      "latencyMs": 11645,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 707,
      "reasoningTokens": 57,
      "cacheReadTokens": 5125,
      "cacheWriteTokens": 2044
    },
    {
      "trialId": "scored:sup-002:openai/gpt-5.6-luna:1",
      "fixtureId": "sup-002",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0009838899999999999,
      "latencyMs": 8034,
      "toolCalls": 4,
      "inputTokens": 803,
      "outputTokens": 294,
      "reasoningTokens": 88,
      "cacheReadTokens": 2332,
      "cacheWriteTokens": 1273
    },
    {
      "trialId": "scored:sup-002:google/gemini-3.8-flash:1",
      "fixtureId": "sup-002",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01175475,
      "latencyMs": 14182,
      "toolCalls": 5,
      "inputTokens": 6083,
      "outputTokens": 332,
      "reasoningTokens": 1586,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-002:anthropic/claude-sonnet-5:1",
      "fixtureId": "sup-002",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0152657,
      "latencyMs": 10669,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 666,
      "reasoningTokens": 62,
      "cacheReadTokens": 4406,
      "cacheWriteTokens": 2837
    },
    {
      "trialId": "scored:ext-003:anthropic/claude-opus-5:1",
      "fixtureId": "ext-003",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.025032250000000002,
      "latencyMs": 8508,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 411,
      "reasoningTokens": 0,
      "cacheReadTokens": 5517,
      "cacheWriteTokens": 1915
    },
    {
      "trialId": "scored:sup-002:openai/gpt-5.6-sol:1",
      "fixtureId": "sup-002",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0083111,
      "latencyMs": 9290,
      "toolCalls": 4,
      "inputTokens": 803,
      "outputTokens": 268,
      "reasoningTokens": 43,
      "cacheReadTokens": 2313,
      "cacheWriteTokens": 1253
    },
    {
      "trialId": "scored:ext-003:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "ext-003",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.004539348,
      "latencyMs": 9926,
      "toolCalls": 3,
      "inputTokens": 1895,
      "outputTokens": 352,
      "reasoningTokens": 130,
      "cacheReadTokens": 2937,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-003:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "ext-003",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000204563,
      "latencyMs": 5152,
      "toolCalls": 3,
      "inputTokens": 714,
      "outputTokens": 467,
      "reasoningTokens": 0,
      "cacheReadTokens": 5632,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-002:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "sup-002",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01014429,
      "latencyMs": 23719,
      "toolCalls": 6,
      "inputTokens": 24,
      "outputTokens": 657,
      "reasoningTokens": 231,
      "cacheReadTokens": 5387,
      "cacheWriteTokens": 1541
    },
    {
      "trialId": "scored:ext-003:google/gemini-3.8-flash:1",
      "fixtureId": "ext-003",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0081165,
      "latencyMs": 10653,
      "toolCalls": 3,
      "inputTokens": 5347,
      "outputTokens": 167,
      "reasoningTokens": 928,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-003:anthropic/claude-sonnet-5:1",
      "fixtureId": "ext-003",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0133272,
      "latencyMs": 11144,
      "toolCalls": 2,
      "inputTokens": 6,
      "outputTokens": 546,
      "reasoningTokens": 0,
      "cacheReadTokens": 4701,
      "cacheWriteTokens": 2766
    },
    {
      "trialId": "scored:ext-003:openai/gpt-5.6-luna:1",
      "fixtureId": "ext-003",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00059967,
      "latencyMs": 7248,
      "toolCalls": 2,
      "inputTokens": 951,
      "outputTokens": 223,
      "reasoningTokens": 28,
      "cacheReadTokens": 3376,
      "cacheWriteTokens": 163
    },
    {
      "trialId": "scored:ext-003:alibaba/qwen3.8-flash:1",
      "fixtureId": "ext-003",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00048758600000000003,
      "latencyMs": 7645,
      "toolCalls": 3,
      "inputTokens": 18,
      "outputTokens": 357,
      "reasoningTokens": 74,
      "cacheReadTokens": 3646,
      "cacheWriteTokens": 1119
    },
    {
      "trialId": "scored:rec-008:google/gemini-3.8-flash:1",
      "fixtureId": "rec-008",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.006738000000000001,
      "latencyMs": 8827,
      "toolCalls": 4,
      "inputTokens": 3544,
      "outputTokens": 248,
      "reasoningTokens": 840,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-003:openai/gpt-5.6-sol:1",
      "fixtureId": "ext-003",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.010625800000000001,
      "latencyMs": 8889,
      "toolCalls": 2,
      "inputTokens": 951,
      "outputTokens": 244,
      "reasoningTokens": 25,
      "cacheReadTokens": 1219,
      "cacheWriteTokens": 2316
    },
    {
      "trialId": "scored:rec-008:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "rec-008",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000195004,
      "latencyMs": 5911,
      "toolCalls": 4,
      "inputTokens": 738,
      "outputTokens": 580,
      "reasoningTokens": 0,
      "cacheReadTokens": 3584,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-008:openai/gpt-5.6-luna:1",
      "fixtureId": "rec-008",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00109107,
      "latencyMs": 7934,
      "toolCalls": 4,
      "inputTokens": 1665,
      "outputTokens": 302,
      "reasoningTokens": 76,
      "cacheReadTokens": 1111,
      "cacheWriteTokens": 1129
    },
    {
      "trialId": "scored:ext-003:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "ext-003",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.007069260000000001,
      "latencyMs": 18266,
      "toolCalls": 3,
      "inputTokens": 18,
      "outputTokens": 368,
      "reasoningTokens": 196,
      "cacheReadTokens": 3628,
      "cacheWriteTokens": 1213
    },
    {
      "trialId": "scored:rec-008:anthropic/claude-sonnet-5:1",
      "fixtureId": "rec-008",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.014103399999999999,
      "latencyMs": 9395,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 586,
      "reasoningTokens": 83,
      "cacheReadTokens": 4107,
      "cacheWriteTokens": 2632
    },
    {
      "trialId": "scored:rec-008:openai/gpt-5.6-sol:1",
      "fixtureId": "rec-008",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0099198,
      "latencyMs": 10623,
      "toolCalls": 4,
      "inputTokens": 1665,
      "outputTokens": 294,
      "reasoningTokens": 57,
      "cacheReadTokens": 1124,
      "cacheWriteTokens": 1142
    },
    {
      "trialId": "scored:rec-008:alibaba/qwen3.8-flash:1",
      "fixtureId": "rec-008",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00077121,
      "latencyMs": 13022,
      "toolCalls": 6,
      "inputTokens": 24,
      "outputTokens": 727,
      "reasoningTokens": 212,
      "cacheReadTokens": 4965,
      "cacheWriteTokens": 1233
    },
    {
      "trialId": "scored:rec-004:openai/gpt-5.6-luna:1",
      "fixtureId": "rec-004",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0006964,
      "latencyMs": 7969,
      "toolCalls": 4,
      "inputTokens": 782,
      "outputTokens": 319,
      "reasoningTokens": 36,
      "cacheReadTokens": 3300,
      "cacheWriteTokens": 192
    },
    {
      "trialId": "scored:rec-008:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "rec-008",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.005862252000000001,
      "latencyMs": 10058,
      "toolCalls": 4,
      "inputTokens": 2829,
      "outputTokens": 404,
      "reasoningTokens": 118,
      "cacheReadTokens": 1383,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-008:anthropic/claude-opus-5:1",
      "fixtureId": "rec-008",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0400585,
      "latencyMs": 14388,
      "toolCalls": 7,
      "inputTokens": 8,
      "outputTokens": 888,
      "reasoningTokens": 50,
      "cacheReadTokens": 7537,
      "cacheWriteTokens": 2048
    },
    {
      "trialId": "scored:rec-004:anthropic/claude-sonnet-5:1",
      "fixtureId": "rec-004",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0073212,
      "latencyMs": 10082,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 581,
      "reasoningTokens": 9,
      "cacheReadTokens": 7046,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-004:anthropic/claude-opus-5:1",
      "fixtureId": "rec-004",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0260205,
      "latencyMs": 11092,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 641,
      "reasoningTokens": 60,
      "cacheReadTokens": 6056,
      "cacheWriteTokens": 870
    },
    {
      "trialId": "scored:rec-004:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "rec-004",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.004689036000000001,
      "latencyMs": 11725,
      "toolCalls": 4,
      "inputTokens": 1846,
      "outputTokens": 417,
      "reasoningTokens": 122,
      "cacheReadTokens": 2679,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-008:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "rec-008",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.010808499999999999,
      "latencyMs": 29449,
      "toolCalls": 7,
      "inputTokens": 30,
      "outputTokens": 750,
      "reasoningTokens": 270,
      "cacheReadTokens": 6550,
      "cacheWriteTokens": 1406
    },
    {
      "trialId": "scored:rec-004:alibaba/qwen3.8-flash:1",
      "fixtureId": "rec-004",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0007363280000000001,
      "latencyMs": 11764,
      "toolCalls": 6,
      "inputTokens": 24,
      "outputTokens": 656,
      "reasoningTokens": 148,
      "cacheReadTokens": 5188,
      "cacheWriteTokens": 1358
    },
    {
      "trialId": "scored:rec-004:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "rec-004",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000202948,
      "latencyMs": 11789,
      "toolCalls": 4,
      "inputTokens": 563,
      "outputTokens": 672,
      "reasoningTokens": 0,
      "cacheReadTokens": 4096,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-004:openai/gpt-5.6-sol:1",
      "fixtureId": "rec-004",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.009114,
      "latencyMs": 11992,
      "toolCalls": 5,
      "inputTokens": 785,
      "outputTokens": 337,
      "reasoningTokens": 32,
      "cacheReadTokens": 3395,
      "cacheWriteTokens": 1270
    },
    {
      "trialId": "scored:ext-009:alibaba/qwen3.8-flash:1",
      "fixtureId": "ext-009",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00041136,
      "latencyMs": 10245,
      "toolCalls": 2,
      "inputTokens": 18,
      "outputTokens": 444,
      "reasoningTokens": 116,
      "cacheReadTokens": 3480,
      "cacheWriteTokens": 448
    },
    {
      "trialId": "scored:ext-009:google/gemini-3.8-flash:1",
      "fixtureId": "ext-009",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00714675,
      "latencyMs": 11684,
      "toolCalls": 3,
      "inputTokens": 4194,
      "outputTokens": 191,
      "reasoningTokens": 876,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-004:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "rec-004",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01108683,
      "latencyMs": 31563,
      "toolCalls": 6,
      "inputTokens": 24,
      "outputTokens": 789,
      "reasoningTokens": 370,
      "cacheReadTokens": 5749,
      "cacheWriteTokens": 1243
    },
    {
      "trialId": "scored:ext-009:openai/gpt-5.6-sol:1",
      "fixtureId": "ext-009",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008862,
      "latencyMs": 7212,
      "toolCalls": 2,
      "inputTokens": 3331,
      "outputTokens": 210,
      "reasoningTokens": 10,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-009:anthropic/claude-sonnet-5:1",
      "fixtureId": "ext-009",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.005971199999999999,
      "latencyMs": 8194,
      "toolCalls": 2,
      "inputTokens": 6,
      "outputTokens": 480,
      "reasoningTokens": 0,
      "cacheReadTokens": 5796,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-009:anthropic/claude-opus-5:1",
      "fixtureId": "ext-009",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.013701,
      "latencyMs": 9019,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 431,
      "reasoningTokens": 0,
      "cacheReadTokens": 5792,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-004:google/gemini-3.8-flash:1",
      "fixtureId": "rec-004",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01057125,
      "latencyMs": 24109,
      "toolCalls": 5,
      "inputTokens": 6325,
      "outputTokens": 453,
      "reasoningTokens": 1101,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-009:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "ext-009",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": false,
      "reasons": [
        "params.contacts: substantive value or membership differs"
      ],
      "cost": 0.00505904,
      "latencyMs": 16600,
      "toolCalls": 3,
      "inputTokens": 18,
      "outputTokens": 443,
      "reasoningTokens": 118,
      "cacheReadTokens": 3512,
      "cacheWriteTokens": 424
    },
    {
      "trialId": "scored:ext-009:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "ext-009",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000175941,
      "latencyMs": 7897,
      "toolCalls": 3,
      "inputTokens": 522,
      "outputTokens": 469,
      "reasoningTokens": 0,
      "cacheReadTokens": 4608,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-006:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "rec-006",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00025094099999999997,
      "latencyMs": 8467,
      "toolCalls": 6,
      "inputTokens": 699,
      "outputTokens": 801,
      "reasoningTokens": 0,
      "cacheReadTokens": 5376,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-009:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "ext-009",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00409904,
      "latencyMs": 9085,
      "toolCalls": 3,
      "inputTokens": 1573,
      "outputTokens": 381,
      "reasoningTokens": 104,
      "cacheReadTokens": 2320,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-009:openai/gpt-5.6-luna:1",
      "fixtureId": "ext-009",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0010744,
      "latencyMs": 9355,
      "toolCalls": 2,
      "inputTokens": 3416,
      "outputTokens": 245,
      "reasoningTokens": 81,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-006:openai/gpt-5.6-sol:1",
      "fixtureId": "rec-006",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0106185,
      "latencyMs": 9759,
      "toolCalls": 4,
      "inputTokens": 2573,
      "outputTokens": 254,
      "reasoningTokens": 34,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 1037
    },
    {
      "trialId": "scored:rec-006:anthropic/claude-sonnet-5:1",
      "fixtureId": "rec-006",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.015021299999999998,
      "latencyMs": 17385,
      "toolCalls": 6,
      "inputTokens": 8,
      "outputTokens": 659,
      "reasoningTokens": 64,
      "cacheReadTokens": 6264,
      "cacheWriteTokens": 2609
    },
    {
      "trialId": "scored:rec-006:openai/gpt-5.6-luna:1",
      "fixtureId": "rec-006",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00100758,
      "latencyMs": 7485,
      "toolCalls": 4,
      "inputTokens": 1554,
      "outputTokens": 253,
      "reasoningTokens": 89,
      "cacheReadTokens": 1044,
      "cacheWriteTokens": 1062
    },
    {
      "trialId": "scored:rec-006:anthropic/claude-opus-5:1",
      "fixtureId": "rec-006",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.025924750000000003,
      "latencyMs": 9201,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 532,
      "reasoningTokens": 38,
      "cacheReadTokens": 4602,
      "cacheWriteTokens": 1495
    },
    {
      "trialId": "scored:rec-006:alibaba/qwen3.8-flash:1",
      "fixtureId": "rec-006",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0007380139999999999,
      "latencyMs": 12088,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 563,
      "reasoningTokens": 318,
      "cacheReadTokens": 4744,
      "cacheWriteTokens": 1221
    },
    {
      "trialId": "scored:rec-006:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "rec-006",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.004961836,
      "latencyMs": 13769,
      "toolCalls": 6,
      "inputTokens": 1804,
      "outputTokens": 431,
      "reasoningTokens": 163,
      "cacheReadTokens": 5189,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-006:google/gemini-3.8-flash:1",
      "fixtureId": "rec-006",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.009709500000000001,
      "latencyMs": 14170,
      "toolCalls": 5,
      "inputTokens": 5331,
      "outputTokens": 340,
      "reasoningTokens": 1183,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-008:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "ext-008",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00019449399999999997,
      "latencyMs": 5590,
      "toolCalls": 4,
      "inputTokens": 572,
      "outputTokens": 542,
      "reasoningTokens": 0,
      "cacheReadTokens": 4864,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-006:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "rec-006",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01066897,
      "latencyMs": 32553,
      "toolCalls": 7,
      "inputTokens": 30,
      "outputTokens": 669,
      "reasoningTokens": 346,
      "cacheReadTokens": 6391,
      "cacheWriteTokens": 1373
    },
    {
      "trialId": "scored:ext-008:openai/gpt-5.6-sol:1",
      "fixtureId": "ext-008",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008682,
      "latencyMs": 8539,
      "toolCalls": 2,
      "inputTokens": 3206,
      "outputTokens": 189,
      "reasoningTokens": 38,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-008:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "ext-008",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.004104584,
      "latencyMs": 8845,
      "toolCalls": 3,
      "inputTokens": 1528,
      "outputTokens": 351,
      "reasoningTokens": 139,
      "cacheReadTokens": 3346,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-008:anthropic/claude-opus-5:1",
      "fixtureId": "ext-008",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.014068250000000001,
      "latencyMs": 8991,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 356,
      "reasoningTokens": 40,
      "cacheReadTokens": 5389,
      "cacheWriteTokens": 231
    },
    {
      "trialId": "scored:ext-008:openai/gpt-5.6-luna:1",
      "fixtureId": "ext-008",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0009881999999999998,
      "latencyMs": 16666,
      "toolCalls": 2,
      "inputTokens": 3231,
      "outputTokens": 197,
      "reasoningTokens": 88,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-008:anthropic/claude-sonnet-5:1",
      "fixtureId": "ext-008",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0069818,
      "latencyMs": 11224,
      "toolCalls": 2,
      "inputTokens": 6,
      "outputTokens": 430,
      "reasoningTokens": 89,
      "cacheReadTokens": 5399,
      "cacheWriteTokens": 280
    },
    {
      "trialId": "scored:ext-008:alibaba/qwen3.8-flash:1",
      "fixtureId": "ext-008",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000799754,
      "latencyMs": 25947,
      "toolCalls": 5,
      "inputTokens": 30,
      "outputTokens": 731,
      "reasoningTokens": 392,
      "cacheReadTokens": 6309,
      "cacheWriteTokens": 831
    },
    {
      "trialId": "scored:ext-008:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "ext-008",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00907528,
      "latencyMs": 27305,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 686,
      "reasoningTokens": 338,
      "cacheReadTokens": 5034,
      "cacheWriteTokens": 811
    },
    {
      "trialId": "scored:rec-010:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "rec-010",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00024835,
      "latencyMs": 7076,
      "toolCalls": 8,
      "inputTokens": 809,
      "outputTokens": 706,
      "reasoningTokens": 0,
      "cacheReadTokens": 5632,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-010:openai/gpt-5.6-luna:1",
      "fixtureId": "rec-010",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00105837,
      "latencyMs": 7789,
      "toolCalls": 6,
      "inputTokens": 1662,
      "outputTokens": 274,
      "reasoningTokens": 66,
      "cacheReadTokens": 1161,
      "cacheWriteTokens": 1179
    },
    {
      "trialId": "scored:rec-010:google/gemini-3.8-flash:1",
      "fixtureId": "rec-010",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00926625,
      "latencyMs": 13135,
      "toolCalls": 9,
      "inputTokens": 5745,
      "outputTokens": 420,
      "reasoningTokens": 902,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-010:anthropic/claude-sonnet-5:1",
      "fixtureId": "rec-010",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0163634,
      "latencyMs": 11267,
      "toolCalls": 6,
      "inputTokens": 6,
      "outputTokens": 767,
      "reasoningTokens": 95,
      "cacheReadTokens": 4157,
      "cacheWriteTokens": 2760
    },
    {
      "trialId": "scored:ext-008:google/gemini-3.8-flash:1",
      "fixtureId": "ext-008",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0095745,
      "latencyMs": 18784,
      "toolCalls": 4,
      "inputTokens": 6356,
      "outputTokens": 254,
      "reasoningTokens": 1028,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-010:anthropic/claude-opus-5:1",
      "fixtureId": "rec-010",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.03347775,
      "latencyMs": 11437,
      "toolCalls": 7,
      "inputTokens": 8,
      "outputTokens": 683,
      "reasoningTokens": 41,
      "cacheReadTokens": 7363,
      "cacheWriteTokens": 1865
    },
    {
      "trialId": "scored:rec-010:alibaba/qwen3.8-flash:1",
      "fixtureId": "rec-010",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0007140599999999999,
      "latencyMs": 11699,
      "toolCalls": 7,
      "inputTokens": 24,
      "outputTokens": 586,
      "reasoningTokens": 220,
      "cacheReadTokens": 4950,
      "cacheWriteTokens": 1261
    },
    {
      "trialId": "scored:rec-010:openai/gpt-5.6-sol:1",
      "fixtureId": "rec-010",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0097435,
      "latencyMs": 8970,
      "toolCalls": 6,
      "inputTokens": 1662,
      "outputTokens": 291,
      "reasoningTokens": 40,
      "cacheReadTokens": 1135,
      "cacheWriteTokens": 1153
    },
    {
      "trialId": "scored:rec-010:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "rec-010",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.006613728,
      "latencyMs": 12341,
      "toolCalls": 7,
      "inputTokens": 2893,
      "outputTokens": 566,
      "reasoningTokens": 105,
      "cacheReadTokens": 3132,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-008:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "sup-008",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000446704,
      "latencyMs": 9096,
      "toolCalls": 8,
      "inputTokens": 1518,
      "outputTokens": 1416,
      "reasoningTokens": 0,
      "cacheReadTokens": 8192,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-008:openai/gpt-5.6-luna:1",
      "fixtureId": "sup-008",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00098459,
      "latencyMs": 10507,
      "toolCalls": 7,
      "inputTokens": 12,
      "outputTokens": 497,
      "reasoningTokens": 143,
      "cacheReadTokens": 6272,
      "cacheWriteTokens": 355
    },
    {
      "trialId": "scored:sup-008:alibaba/qwen3.8-flash:1",
      "fixtureId": "sup-008",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0015221260000000002,
      "latencyMs": 26826,
      "toolCalls": 8,
      "inputTokens": 54,
      "outputTokens": 1190,
      "reasoningTokens": 563,
      "cacheReadTokens": 17261,
      "cacheWriteTokens": 2067
    },
    {
      "trialId": "scored:rec-010:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "rec-010",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01260065,
      "latencyMs": 32396,
      "toolCalls": 11,
      "inputTokens": 24,
      "outputTokens": 937,
      "reasoningTokens": 329,
      "cacheReadTokens": 5495,
      "cacheWriteTokens": 1609
    },
    {
      "trialId": "scored:sup-008:openai/gpt-5.6-sol:1",
      "fixtureId": "sup-008",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.006832,
      "latencyMs": 8283,
      "toolCalls": 7,
      "inputTokens": 9,
      "outputTokens": 313,
      "reasoningTokens": 82,
      "cacheReadTokens": 3720,
      "cacheWriteTokens": 848
    },
    {
      "trialId": "scored:sup-008:google/gemini-3.8-flash:1",
      "fixtureId": "sup-008",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01864575,
      "latencyMs": 22321,
      "toolCalls": 7,
      "inputTokens": 8801,
      "outputTokens": 421,
      "reasoningTokens": 2791,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-008:anthropic/claude-opus-5:1",
      "fixtureId": "sup-008",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.053406249999999995,
      "latencyMs": 22969,
      "toolCalls": 7,
      "inputTokens": 6,
      "outputTokens": 1361,
      "reasoningTokens": 116,
      "cacheReadTokens": 8090,
      "cacheWriteTokens": 1985
    },
    {
      "trialId": "scored:rec-005:openai/gpt-5.6-luna:1",
      "fixtureId": "rec-005",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0009172000000000001,
      "latencyMs": 8613,
      "toolCalls": 3,
      "inputTokens": 3062,
      "outputTokens": 192,
      "reasoningTokens": 62,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-005:openai/gpt-5.6-sol:1",
      "fixtureId": "rec-005",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.010392000000000002,
      "latencyMs": 9515,
      "toolCalls": 4,
      "inputTokens": 3966,
      "outputTokens": 208,
      "reasoningTokens": 38,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-005:alibaba/qwen3.8-flash:1",
      "fixtureId": "rec-005",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000507224,
      "latencyMs": 9916,
      "toolCalls": 4,
      "inputTokens": 24,
      "outputTokens": 373,
      "reasoningTokens": 199,
      "cacheReadTokens": 4284,
      "cacheWriteTokens": 830
    },
    {
      "trialId": "scored:sup-008:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "sup-008",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.010969288,
      "latencyMs": 24046,
      "toolCalls": 8,
      "inputTokens": 3446,
      "outputTokens": 821,
      "reasoningTokens": 741,
      "cacheReadTokens": 5342,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-005:anthropic/claude-opus-5:1",
      "fixtureId": "rec-005",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.02144825,
      "latencyMs": 9066,
      "toolCalls": 4,
      "inputTokens": 8,
      "outputTokens": 425,
      "reasoningTokens": 0,
      "cacheReadTokens": 6129,
      "cacheWriteTokens": 1235
    },
    {
      "trialId": "scored:sup-008:anthropic/claude-sonnet-5:1",
      "fixtureId": "sup-008",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.030596200000000004,
      "latencyMs": 20724,
      "toolCalls": 7,
      "inputTokens": 6,
      "outputTokens": 1271,
      "reasoningTokens": 498,
      "cacheReadTokens": 5896,
      "cacheWriteTokens": 4686
    },
    {
      "trialId": "scored:rec-005:anthropic/claude-sonnet-5:1",
      "fixtureId": "rec-005",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.010343100000000001,
      "latencyMs": 9866,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 374,
      "reasoningTokens": 54,
      "cacheReadTokens": 3518,
      "cacheWriteTokens": 2139
    },
    {
      "trialId": "scored:rec-005:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "rec-005",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00786976,
      "latencyMs": 22243,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 532,
      "reasoningTokens": 198,
      "cacheReadTokens": 4378,
      "cacheWriteTokens": 1079
    },
    {
      "trialId": "scored:rec-005:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "rec-005",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00600072,
      "latencyMs": 13843,
      "toolCalls": 3,
      "inputTokens": 3502,
      "outputTokens": 302,
      "reasoningTokens": 46,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-005:google/gemini-3.8-flash:1",
      "fixtureId": "rec-005",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.006581249999999999,
      "latencyMs": 10547,
      "toolCalls": 5,
      "inputTokens": 4845,
      "outputTokens": 261,
      "reasoningTokens": 525,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-008:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "sup-008",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.02034005,
      "latencyMs": 52268,
      "toolCalls": 8,
      "inputTokens": 24,
      "outputTokens": 1399,
      "reasoningTokens": 767,
      "cacheReadTokens": 7065,
      "cacheWriteTokens": 2438
    },
    {
      "trialId": "scored:rec-005:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "rec-005",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000171537,
      "latencyMs": 7751,
      "toolCalls": 4,
      "inputTokens": 632,
      "outputTokens": 409,
      "reasoningTokens": 0,
      "cacheReadTokens": 4352,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-002:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "ext-002",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.006680036,
      "latencyMs": 10658,
      "toolCalls": 4,
      "inputTokens": 3107,
      "outputTokens": 486,
      "reasoningTokens": 134,
      "cacheReadTokens": 2809,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-002:alibaba/qwen3.8-flash:1",
      "fixtureId": "ext-002",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000685366,
      "latencyMs": 11146,
      "toolCalls": 3,
      "inputTokens": 18,
      "outputTokens": 557,
      "reasoningTokens": 224,
      "cacheReadTokens": 3451,
      "cacheWriteTokens": 1301
    },
    {
      "trialId": "scored:ext-002:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "ext-002",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000268446,
      "latencyMs": 7530,
      "toolCalls": 4,
      "inputTokens": 1032,
      "outputTokens": 750,
      "reasoningTokens": 0,
      "cacheReadTokens": 5376,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-002:openai/gpt-5.6-luna:1",
      "fixtureId": "ext-002",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0007338099999999999,
      "latencyMs": 10683,
      "toolCalls": 3,
      "inputTokens": 785,
      "outputTokens": 326,
      "reasoningTokens": 57,
      "cacheReadTokens": 3348,
      "cacheWriteTokens": 201
    },
    {
      "trialId": "scored:ext-002:anthropic/claude-opus-5:1",
      "fixtureId": "ext-002",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.027914249999999998,
      "latencyMs": 12226,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 552,
      "reasoningTokens": 20,
      "cacheReadTokens": 4981,
      "cacheWriteTokens": 1775
    },
    {
      "trialId": "scored:ext-002:openai/gpt-5.6-sol:1",
      "fixtureId": "ext-002",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008645100000000001,
      "latencyMs": 9099,
      "toolCalls": 3,
      "inputTokens": 785,
      "outputTokens": 324,
      "reasoningTokens": 25,
      "cacheReadTokens": 2288,
      "cacheWriteTokens": 1251
    },
    {
      "trialId": "scored:ext-002:anthropic/claude-sonnet-5:1",
      "fixtureId": "ext-002",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0146307,
      "latencyMs": 12405,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 710,
      "reasoningTokens": 0,
      "cacheReadTokens": 4256,
      "cacheWriteTokens": 2667
    },
    {
      "trialId": "scored:rec-003:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "rec-003",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.006915831999999999,
      "latencyMs": 12483,
      "toolCalls": 6,
      "inputTokens": 3248,
      "outputTokens": 508,
      "reasoningTokens": 125,
      "cacheReadTokens": 2768,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-002:google/gemini-3.8-flash:1",
      "fixtureId": "ext-002",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00927675,
      "latencyMs": 12626,
      "toolCalls": 4,
      "inputTokens": 5549,
      "outputTokens": 413,
      "reasoningTokens": 951,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-003:openai/gpt-5.6-sol:1",
      "fixtureId": "rec-003",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0104088,
      "latencyMs": 14090,
      "toolCalls": 7,
      "inputTokens": 759,
      "outputTokens": 370,
      "reasoningTokens": 76,
      "cacheReadTokens": 4779,
      "cacheWriteTokens": 1390
    },
    {
      "trialId": "scored:rec-003:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "rec-003",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000264833,
      "latencyMs": 6623,
      "toolCalls": 7,
      "inputTokens": 1067,
      "outputTokens": 709,
      "reasoningTokens": 0,
      "cacheReadTokens": 5376,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-002:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "ext-002",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0114472,
      "latencyMs": 34603,
      "toolCalls": 5,
      "inputTokens": 30,
      "outputTokens": 666,
      "reasoningTokens": 404,
      "cacheReadTokens": 6910,
      "cacheWriteTokens": 1517
    },
    {
      "trialId": "scored:rec-003:alibaba/qwen3.8-flash:1",
      "fixtureId": "rec-003",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000955208,
      "latencyMs": 15936,
      "toolCalls": 7,
      "inputTokens": 24,
      "outputTokens": 776,
      "reasoningTokens": 372,
      "cacheReadTokens": 5388,
      "cacheWriteTokens": 1628
    },
    {
      "trialId": "scored:rec-003:openai/gpt-5.6-luna:1",
      "fixtureId": "rec-003",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0010758999999999999,
      "latencyMs": 8307,
      "toolCalls": 5,
      "inputTokens": 753,
      "outputTokens": 375,
      "reasoningTokens": 89,
      "cacheReadTokens": 2325,
      "cacheWriteTokens": 1288
    },
    {
      "trialId": "scored:rec-003:anthropic/claude-sonnet-5:1",
      "fixtureId": "rec-003",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0163047,
      "latencyMs": 11266,
      "toolCalls": 5,
      "inputTokens": 6,
      "outputTokens": 698,
      "reasoningTokens": 115,
      "cacheReadTokens": 4351,
      "cacheWriteTokens": 2917
    },
    {
      "trialId": "scored:sup-009:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "sup-009",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000334293,
      "latencyMs": 5986,
      "toolCalls": 6,
      "inputTokens": 765,
      "outputTokens": 1313,
      "reasoningTokens": 0,
      "cacheReadTokens": 5376,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-009:openai/gpt-5.6-sol:1",
      "fixtureId": "sup-009",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.012508,
      "latencyMs": 7229,
      "toolCalls": 6,
      "inputTokens": 889,
      "outputTokens": 279,
      "reasoningTokens": 43,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 3004
    },
    {
      "trialId": "scored:rec-003:anthropic/claude-opus-5:1",
      "fixtureId": "rec-003",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.03288175,
      "latencyMs": 11468,
      "toolCalls": 5,
      "inputTokens": 6,
      "outputTokens": 698,
      "reasoningTokens": 34,
      "cacheReadTokens": 5066,
      "cacheWriteTokens": 1923
    },
    {
      "trialId": "scored:sup-009:anthropic/claude-sonnet-5:1",
      "fixtureId": "sup-009",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0187311,
      "latencyMs": 11688,
      "toolCalls": 6,
      "inputTokens": 6,
      "outputTokens": 756,
      "reasoningTokens": 125,
      "cacheReadTokens": 5183,
      "cacheWriteTokens": 3549
    },
    {
      "trialId": "scored:rec-003:google/gemini-3.8-flash:1",
      "fixtureId": "rec-003",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.010358250000000001,
      "latencyMs": 12913,
      "toolCalls": 8,
      "inputTokens": 7336,
      "outputTokens": 562,
      "reasoningTokens": 733,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-009:alibaba/qwen3.8-flash:1",
      "fixtureId": "sup-009",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.001178758,
      "latencyMs": 17186,
      "toolCalls": 8,
      "inputTokens": 24,
      "outputTokens": 989,
      "reasoningTokens": 376,
      "cacheReadTokens": 6423,
      "cacheWriteTokens": 2153
    },
    {
      "trialId": "scored:rec-003:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "rec-003",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0132309,
      "latencyMs": 33681,
      "toolCalls": 8,
      "inputTokens": 24,
      "outputTokens": 926,
      "reasoningTokens": 350,
      "cacheReadTokens": 5320,
      "cacheWriteTokens": 1849
    },
    {
      "trialId": "scored:sup-009:openai/gpt-5.6-luna:1",
      "fixtureId": "sup-009",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00122807,
      "latencyMs": 8551,
      "toolCalls": 6,
      "inputTokens": 892,
      "outputTokens": 375,
      "reasoningTokens": 107,
      "cacheReadTokens": 3026,
      "cacheWriteTokens": 1643
    },
    {
      "trialId": "scored:sup-010:openai/gpt-5.6-luna:1",
      "fixtureId": "sup-010",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0010907299999999998,
      "latencyMs": 8491,
      "toolCalls": 4,
      "inputTokens": 904,
      "outputTokens": 341,
      "reasoningTokens": 69,
      "cacheReadTokens": 2684,
      "cacheWriteTokens": 1457
    },
    {
      "trialId": "scored:sup-009:anthropic/claude-opus-5:1",
      "fixtureId": "sup-009",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.04599875,
      "latencyMs": 15888,
      "toolCalls": 6,
      "inputTokens": 6,
      "outputTokens": 937,
      "reasoningTokens": 99,
      "cacheReadTokens": 5900,
      "cacheWriteTokens": 2739
    },
    {
      "trialId": "scored:sup-010:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "sup-010",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000232853,
      "latencyMs": 7417,
      "toolCalls": 4,
      "inputTokens": 683,
      "outputTokens": 761,
      "reasoningTokens": 0,
      "cacheReadTokens": 4608,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-009:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "sup-009",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008508808,
      "latencyMs": 17243,
      "toolCalls": 7,
      "inputTokens": 3883,
      "outputTokens": 506,
      "reasoningTokens": 310,
      "cacheReadTokens": 3452,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-009:google/gemini-3.8-flash:1",
      "fixtureId": "sup-009",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.013728,
      "latencyMs": 16200,
      "toolCalls": 7,
      "inputTokens": 7484,
      "outputTokens": 439,
      "reasoningTokens": 1725,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-010:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "sup-010",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008018955999999999,
      "latencyMs": 17338,
      "toolCalls": 5,
      "inputTokens": 3560,
      "outputTokens": 575,
      "reasoningTokens": 228,
      "cacheReadTokens": 3179,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-010:alibaba/qwen3.8-flash:1",
      "fixtureId": "sup-010",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000801974,
      "latencyMs": 18770,
      "toolCalls": 5,
      "inputTokens": 18,
      "outputTokens": 317,
      "reasoningTokens": 484,
      "cacheReadTokens": 3889,
      "cacheWriteTokens": 1802
    },
    {
      "trialId": "scored:sup-010:openai/gpt-5.6-sol:1",
      "fixtureId": "sup-010",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0093133,
      "latencyMs": 8557,
      "toolCalls": 4,
      "inputTokens": 904,
      "outputTokens": 306,
      "reasoningTokens": 32,
      "cacheReadTokens": 2664,
      "cacheWriteTokens": 1437
    },
    {
      "trialId": "scored:sup-010:anthropic/claude-opus-5:1",
      "fixtureId": "sup-010",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.04045025000000001,
      "latencyMs": 12382,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 838,
      "reasoningTokens": 79,
      "cacheReadTokens": 5578,
      "cacheWriteTokens": 2353
    },
    {
      "trialId": "scored:sup-010:anthropic/claude-sonnet-5:1",
      "fixtureId": "sup-010",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.017974,
      "latencyMs": 15331,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 793,
      "reasoningTokens": 93,
      "cacheReadTokens": 4860,
      "cacheWriteTokens": 3252
    },
    {
      "trialId": "scored:rec-002:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "rec-002",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.004038452,
      "latencyMs": 6490,
      "toolCalls": 4,
      "inputTokens": 1607,
      "outputTokens": 394,
      "reasoningTokens": 63,
      "cacheReadTokens": 2443,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-010:google/gemini-3.8-flash:1",
      "fixtureId": "sup-010",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01137525,
      "latencyMs": 17100,
      "toolCalls": 5,
      "inputTokens": 6462,
      "outputTokens": 352,
      "reasoningTokens": 1389,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-002:openai/gpt-5.6-luna:1",
      "fixtureId": "rec-002",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00099717,
      "latencyMs": 7795,
      "toolCalls": 4,
      "inputTokens": 1605,
      "outputTokens": 246,
      "reasoningTokens": 75,
      "cacheReadTokens": 1061,
      "cacheWriteTokens": 1079
    },
    {
      "trialId": "scored:rec-002:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "rec-002",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000219953,
      "latencyMs": 4836,
      "toolCalls": 5,
      "inputTokens": 1108,
      "outputTokens": 489,
      "reasoningTokens": 0,
      "cacheReadTokens": 4352,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-002:anthropic/claude-sonnet-5:1",
      "fixtureId": "rec-002",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.012101299999999999,
      "latencyMs": 10219,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 449,
      "reasoningTokens": 55,
      "cacheReadTokens": 3984,
      "cacheWriteTokens": 2501
    },
    {
      "trialId": "scored:rec-002:google/gemini-3.8-flash:1",
      "fixtureId": "rec-002",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00940425,
      "latencyMs": 11549,
      "toolCalls": 5,
      "inputTokens": 5834,
      "outputTokens": 231,
      "reasoningTokens": 1110,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-002:openai/gpt-5.6-sol:1",
      "fixtureId": "rec-002",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0090221,
      "latencyMs": 8611,
      "toolCalls": 4,
      "inputTokens": 1602,
      "outputTokens": 260,
      "reasoningTokens": 33,
      "cacheReadTokens": 1053,
      "cacheWriteTokens": 1071
    },
    {
      "trialId": "scored:sup-009:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "sup-009",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.02002002,
      "latencyMs": 54884,
      "toolCalls": 9,
      "inputTokens": 36,
      "outputTokens": 1643,
      "reasoningTokens": 404,
      "cacheReadTokens": 10006,
      "cacheWriteTokens": 2386
    },
    {
      "trialId": "scored:rec-002:anthropic/claude-opus-5:1",
      "fixtureId": "rec-002",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.035902750000000004,
      "latencyMs": 12699,
      "toolCalls": 6,
      "inputTokens": 8,
      "outputTokens": 752,
      "reasoningTokens": 62,
      "cacheReadTokens": 7113,
      "cacheWriteTokens": 1913
    },
    {
      "trialId": "scored:rec-002:alibaba/qwen3.8-flash:1",
      "fixtureId": "rec-002",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0006828100000000001,
      "latencyMs": 13556,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 544,
      "reasoningTokens": 223,
      "cacheReadTokens": 4855,
      "cacheWriteTokens": 1204
    },
    {
      "trialId": "scored:ext-006:anthropic/claude-sonnet-5:1",
      "fixtureId": "ext-006",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.013995599999999999,
      "latencyMs": 9129,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 591,
      "reasoningTokens": 39,
      "cacheReadTokens": 4293,
      "cacheWriteTokens": 2730
    },
    {
      "trialId": "scored:ext-006:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "ext-006",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00810348,
      "latencyMs": 9946,
      "toolCalls": 4,
      "inputTokens": 4498,
      "outputTokens": 438,
      "reasoningTokens": 109,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-006:anthropic/claude-opus-5:1",
      "fixtureId": "ext-006",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.03479275,
      "latencyMs": 12135,
      "toolCalls": 5,
      "inputTokens": 8,
      "outputTokens": 697,
      "reasoningTokens": 63,
      "cacheReadTokens": 7518,
      "cacheWriteTokens": 1919
    },
    {
      "trialId": "scored:rec-002:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "rec-002",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01017846,
      "latencyMs": 28784,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 678,
      "reasoningTokens": 320,
      "cacheReadTokens": 4838,
      "cacheWriteTokens": 1328
    },
    {
      "trialId": "scored:ext-006:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "ext-006",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000255213,
      "latencyMs": 5554,
      "toolCalls": 6,
      "inputTokens": 853,
      "outputTokens": 729,
      "reasoningTokens": 0,
      "cacheReadTokens": 5632,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-010:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "sup-010",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.015440599999999999,
      "latencyMs": 39243,
      "toolCalls": 7,
      "inputTokens": 24,
      "outputTokens": 933,
      "reasoningTokens": 511,
      "cacheReadTokens": 6080,
      "cacheWriteTokens": 2278
    },
    {
      "trialId": "scored:ext-006:openai/gpt-5.6-luna:1",
      "fixtureId": "ext-006",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0007203800000000001,
      "latencyMs": 7384,
      "toolCalls": 4,
      "inputTokens": 751,
      "outputTokens": 310,
      "reasoningTokens": 70,
      "cacheReadTokens": 3334,
      "cacheWriteTokens": 190
    },
    {
      "trialId": "scored:ext-006:alibaba/qwen3.8-flash:1",
      "fixtureId": "ext-006",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0009594079999999999,
      "latencyMs": 14411,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 777,
      "reasoningTokens": 359,
      "cacheReadTokens": 5378,
      "cacheWriteTokens": 1678
    },
    {
      "trialId": "scored:ext-006:openai/gpt-5.6-sol:1",
      "fixtureId": "ext-006",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0060258,
      "latencyMs": 9185,
      "toolCalls": 4,
      "inputTokens": 751,
      "outputTokens": 306,
      "reasoningTokens": 34,
      "cacheReadTokens": 3319,
      "cacheWriteTokens": 184
    },
    {
      "trialId": "scored:sup-001:google/gemini-3.8-flash:1",
      "fixtureId": "sup-001",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00892575,
      "latencyMs": 10850,
      "toolCalls": 6,
      "inputTokens": 6246,
      "outputTokens": 289,
      "reasoningTokens": 842,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-006:google/gemini-3.8-flash:1",
      "fixtureId": "ext-006",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0100935,
      "latencyMs": 12589,
      "toolCalls": 6,
      "inputTokens": 7033,
      "outputTokens": 468,
      "reasoningTokens": 817,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-001:openai/gpt-5.6-luna:1",
      "fixtureId": "sup-001",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00107996,
      "latencyMs": 8769,
      "toolCalls": 5,
      "inputTokens": 750,
      "outputTokens": 379,
      "reasoningTokens": 71,
      "cacheReadTokens": 2473,
      "cacheWriteTokens": 1362
    },
    {
      "trialId": "scored:sup-001:openai/gpt-5.6-sol:1",
      "fixtureId": "sup-001",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.009218,
      "latencyMs": 9779,
      "toolCalls": 5,
      "inputTokens": 750,
      "outputTokens": 355,
      "reasoningTokens": 35,
      "cacheReadTokens": 2440,
      "cacheWriteTokens": 1332
    },
    {
      "trialId": "scored:sup-001:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "sup-001",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.006768652,
      "latencyMs": 9816,
      "toolCalls": 6,
      "inputTokens": 3188,
      "outputTokens": 496,
      "reasoningTokens": 132,
      "cacheReadTokens": 1673,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-001:anthropic/claude-sonnet-5:1",
      "fixtureId": "sup-001",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0147241,
      "latencyMs": 14116,
      "toolCalls": 5,
      "inputTokens": 6,
      "outputTokens": 573,
      "reasoningTokens": 82,
      "cacheReadTokens": 4398,
      "cacheWriteTokens": 2913
    },
    {
      "trialId": "scored:sup-001:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "sup-001",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000259301,
      "latencyMs": 8068,
      "toolCalls": 5,
      "inputTokens": 1100,
      "outputTokens": 797,
      "reasoningTokens": 0,
      "cacheReadTokens": 3840,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-001:anthropic/claude-opus-5:1",
      "fixtureId": "sup-001",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.03561975,
      "latencyMs": 12300,
      "toolCalls": 5,
      "inputTokens": 6,
      "outputTokens": 711,
      "reasoningTokens": 72,
      "cacheReadTokens": 5117,
      "cacheWriteTokens": 2153
    },
    {
      "trialId": "scored:ext-006:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "ext-006",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01228768,
      "latencyMs": 33898,
      "toolCalls": 8,
      "inputTokens": 30,
      "outputTokens": 808,
      "reasoningTokens": 313,
      "cacheReadTokens": 7554,
      "cacheWriteTokens": 1687
    },
    {
      "trialId": "scored:ext-005:alibaba/qwen3.8-flash:1",
      "fixtureId": "ext-005",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000611186,
      "latencyMs": 13488,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 503,
      "reasoningTokens": 192,
      "cacheReadTokens": 4706,
      "cacheWriteTokens": 1027
    },
    {
      "trialId": "scored:ext-005:anthropic/claude-opus-5:1",
      "fixtureId": "ext-005",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.03083375,
      "latencyMs": 13647,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 677,
      "reasoningTokens": 61,
      "cacheReadTokens": 4595,
      "cacheWriteTokens": 1609
    },
    {
      "trialId": "scored:sup-001:alibaba/qwen3.8-flash:1",
      "fixtureId": "sup-001",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0012942460000000002,
      "latencyMs": 23628,
      "toolCalls": 8,
      "inputTokens": 54,
      "outputTokens": 847,
      "reasoningTokens": 522,
      "cacheReadTokens": 14211,
      "cacheWriteTokens": 2074
    },
    {
      "trialId": "scored:ext-005:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "ext-005",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.006243116,
      "latencyMs": 15717,
      "toolCalls": 5,
      "inputTokens": 2853,
      "outputTokens": 369,
      "reasoningTokens": 241,
      "cacheReadTokens": 1399,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-005:openai/gpt-5.6-luna:1",
      "fixtureId": "ext-005",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0009900900000000001,
      "latencyMs": 9081,
      "toolCalls": 4,
      "inputTokens": 1560,
      "outputTokens": 250,
      "reasoningTokens": 78,
      "cacheReadTokens": 1037,
      "cacheWriteTokens": 1055
    },
    {
      "trialId": "scored:ext-005:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "ext-005",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00024352,
      "latencyMs": 7362,
      "toolCalls": 5,
      "inputTokens": 700,
      "outputTokens": 752,
      "reasoningTokens": 0,
      "cacheReadTokens": 5376,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-005:openai/gpt-5.6-sol:1",
      "fixtureId": "ext-005",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.010625500000000001,
      "latencyMs": 8541,
      "toolCalls": 4,
      "inputTokens": 2574,
      "outputTokens": 252,
      "reasoningTokens": 37,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 1035
    },
    {
      "trialId": "scored:ext-005:anthropic/claude-sonnet-5:1",
      "fixtureId": "ext-005",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0137472,
      "latencyMs": 10518,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 597,
      "reasoningTokens": 85,
      "cacheReadTokens": 3876,
      "cacheWriteTokens": 2456
    },
    {
      "trialId": "scored:ext-005:google/gemini-3.8-flash:1",
      "fixtureId": "ext-005",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.009558,
      "latencyMs": 14043,
      "toolCalls": 5,
      "inputTokens": 5624,
      "outputTokens": 277,
      "reasoningTokens": 1147,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-005:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "ext-005",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00805672,
      "latencyMs": 22110,
      "toolCalls": 6,
      "inputTokens": 24,
      "outputTokens": 538,
      "reasoningTokens": 187,
      "cacheReadTokens": 4816,
      "cacheWriteTokens": 1136
    },
    {
      "trialId": "scored:sup-001:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "sup-001",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01260124,
      "latencyMs": 35844,
      "toolCalls": 7,
      "inputTokens": 24,
      "outputTokens": 710,
      "reasoningTokens": 454,
      "cacheReadTokens": 5672,
      "cacheWriteTokens": 1842
    },
    {
      "trialId": "scored:rec-009:openai/gpt-5.6-luna:1",
      "fixtureId": "rec-009",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00112848,
      "latencyMs": 9881,
      "toolCalls": 4,
      "inputTokens": 885,
      "outputTokens": 337,
      "reasoningTokens": 136,
      "cacheReadTokens": 2419,
      "cacheWriteTokens": 1342
    },
    {
      "trialId": "scored:rec-009:alibaba/qwen3.8-flash:1",
      "fixtureId": "rec-009",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.001027418,
      "latencyMs": 15530,
      "toolCalls": 4,
      "inputTokens": 18,
      "outputTokens": 833,
      "reasoningTokens": 454,
      "cacheReadTokens": 3578,
      "cacheWriteTokens": 1812
    },
    {
      "trialId": "scored:rec-009:anthropic/claude-opus-5:1",
      "fixtureId": "rec-009",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.053601750000000004,
      "latencyMs": 18711,
      "toolCalls": 5,
      "inputTokens": 8,
      "outputTokens": 1124,
      "reasoningTokens": 212,
      "cacheReadTokens": 8186,
      "cacheWriteTokens": 2571
    },
    {
      "trialId": "scored:rec-009:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "rec-009",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00773564,
      "latencyMs": 17803,
      "toolCalls": 4,
      "inputTokens": 3298,
      "outputTokens": 576,
      "reasoningTokens": 261,
      "cacheReadTokens": 1540,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-009:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "rec-009",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000318228,
      "latencyMs": 9224,
      "toolCalls": 6,
      "inputTokens": 627,
      "outputTokens": 1136,
      "reasoningTokens": 0,
      "cacheReadTokens": 6912,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-009:openai/gpt-5.6-sol:1",
      "fixtureId": "rec-009",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0095346,
      "latencyMs": 9788,
      "toolCalls": 4,
      "inputTokens": 885,
      "outputTokens": 350,
      "reasoningTokens": 55,
      "cacheReadTokens": 2373,
      "cacheWriteTokens": 1296
    },
    {
      "trialId": "scored:rec-009:anthropic/claude-sonnet-5:1",
      "fixtureId": "rec-009",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0242482,
      "latencyMs": 19184,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 879,
      "reasoningTokens": 586,
      "cacheReadTokens": 4706,
      "cacheWriteTokens": 3458
    },
    {
      "trialId": "scored:sup-004:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "sup-004",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000232021,
      "latencyMs": 6103,
      "toolCalls": 5,
      "inputTokens": 840,
      "outputTokens": 701,
      "reasoningTokens": 0,
      "cacheReadTokens": 4352,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-009:google/gemini-3.8-flash:1",
      "fixtureId": "rec-009",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.023715,
      "latencyMs": 29136,
      "toolCalls": 5,
      "inputTokens": 10450,
      "outputTokens": 521,
      "reasoningTokens": 3713,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-004:anthropic/claude-opus-5:1",
      "fixtureId": "sup-004",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.03801425,
      "latencyMs": 12577,
      "toolCalls": 5,
      "inputTokens": 6,
      "outputTokens": 767,
      "reasoningTokens": 67,
      "cacheReadTokens": 5456,
      "cacheWriteTokens": 2305
    },
    {
      "trialId": "scored:sup-004:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "sup-004",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.005469552,
      "latencyMs": 15480,
      "toolCalls": 5,
      "inputTokens": 2116,
      "outputTokens": 473,
      "reasoningTokens": 170,
      "cacheReadTokens": 2958,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-004:google/gemini-3.8-flash:1",
      "fixtureId": "sup-004",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00871725,
      "latencyMs": 13248,
      "toolCalls": 5,
      "inputTokens": 4713,
      "outputTokens": 335,
      "reasoningTokens": 1047,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-004:anthropic/claude-sonnet-5:1",
      "fixtureId": "sup-004",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0179862,
      "latencyMs": 17052,
      "toolCalls": 6,
      "inputTokens": 8,
      "outputTokens": 661,
      "reasoningTokens": 161,
      "cacheReadTokens": 7651,
      "cacheWriteTokens": 3288
    },
    {
      "trialId": "scored:sup-004:openai/gpt-5.6-sol:1",
      "fixtureId": "sup-004",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0072513000000000005,
      "latencyMs": 9727,
      "toolCalls": 5,
      "inputTokens": 844,
      "outputTokens": 356,
      "reasoningTokens": 62,
      "cacheReadTokens": 3929,
      "cacheWriteTokens": 239
    },
    {
      "trialId": "scored:sup-004:openai/gpt-5.6-luna:1",
      "fixtureId": "sup-004",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00122309,
      "latencyMs": 9319,
      "toolCalls": 5,
      "inputTokens": 844,
      "outputTokens": 377,
      "reasoningTokens": 139,
      "cacheReadTokens": 2742,
      "cacheWriteTokens": 1521
    },
    {
      "trialId": "scored:rec-007:anthropic/claude-sonnet-5:1",
      "fixtureId": "rec-007",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0138361,
      "latencyMs": 9836,
      "toolCalls": 5,
      "inputTokens": 6,
      "outputTokens": 591,
      "reasoningTokens": 34,
      "cacheReadTokens": 4208,
      "cacheWriteTokens": 2693
    },
    {
      "trialId": "scored:rec-007:alibaba/qwen3.8-flash:1",
      "fixtureId": "rec-007",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0007412740000000001,
      "latencyMs": 16018,
      "toolCalls": 7,
      "inputTokens": 30,
      "outputTokens": 453,
      "reasoningTokens": 278,
      "cacheReadTokens": 6494,
      "cacheWriteTokens": 1445
    },
    {
      "trialId": "scored:rec-007:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "rec-007",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00025879599999999994,
      "latencyMs": 7611,
      "toolCalls": 6,
      "inputTokens": 1077,
      "outputTokens": 688,
      "reasoningTokens": 0,
      "cacheReadTokens": 5120,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-007:openai/gpt-5.6-sol:1",
      "fixtureId": "rec-007",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0110126,
      "latencyMs": 10658,
      "toolCalls": 6,
      "inputTokens": 1748,
      "outputTokens": 341,
      "reasoningTokens": 49,
      "cacheReadTokens": 2333,
      "cacheWriteTokens": 1260
    },
    {
      "trialId": "scored:rec-007:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "rec-007",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.005972516000000001,
      "latencyMs": 15266,
      "toolCalls": 5,
      "inputTokens": 2868,
      "outputTokens": 489,
      "reasoningTokens": 47,
      "cacheReadTokens": 1459,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-004:alibaba/qwen3.8-flash:1",
      "fixtureId": "sup-004",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.001013702,
      "latencyMs": 26253,
      "toolCalls": 6,
      "inputTokens": 42,
      "outputTokens": 757,
      "reasoningTokens": 268,
      "cacheReadTokens": 10552,
      "cacheWriteTokens": 1782
    },
    {
      "trialId": "scored:sup-004:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "sup-004",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.015052109999999999,
      "latencyMs": 38289,
      "toolCalls": 7,
      "inputTokens": 24,
      "outputTokens": 952,
      "reasoningTokens": 444,
      "cacheReadTokens": 6283,
      "cacheWriteTokens": 2224
    },
    {
      "trialId": "scored:rec-007:anthropic/claude-opus-5:1",
      "fixtureId": "rec-007",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.032917,
      "latencyMs": 12320,
      "toolCalls": 6,
      "inputTokens": 8,
      "outputTokens": 655,
      "reasoningTokens": 41,
      "cacheReadTokens": 7354,
      "cacheWriteTokens": 1888
    },
    {
      "trialId": "scored:rec-007:google/gemini-3.8-flash:1",
      "fixtureId": "rec-007",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01154175,
      "latencyMs": 14324,
      "toolCalls": 6,
      "inputTokens": 5874,
      "outputTokens": 481,
      "reasoningTokens": 1422,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-009:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "rec-009",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01707021,
      "latencyMs": 57886,
      "toolCalls": 6,
      "inputTokens": 30,
      "outputTokens": 843,
      "reasoningTokens": 962,
      "cacheReadTokens": 7663,
      "cacheWriteTokens": 1951
    },
    {
      "trialId": "scored:rec-007:openai/gpt-5.6-luna:1",
      "fixtureId": "rec-007",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0011300699999999997,
      "latencyMs": 7442,
      "toolCalls": 5,
      "inputTokens": 1749,
      "outputTokens": 312,
      "reasoningTokens": 62,
      "cacheReadTokens": 1211,
      "cacheWriteTokens": 1229
    },
    {
      "trialId": "scored:sup-007:openai/gpt-5.6-luna:1",
      "fixtureId": "sup-007",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": false,
      "reasons": [
        "params.monthlyFeeCents: substantive value or membership differs",
        "params.creditAmountCents: substantive value or membership differs"
      ],
      "cost": 0.00114587,
      "latencyMs": 9696,
      "toolCalls": 6,
      "inputTokens": 795,
      "outputTokens": 271,
      "reasoningTokens": 184,
      "cacheReadTokens": 2856,
      "cacheWriteTokens": 1535
    },
    {
      "trialId": "scored:sup-007:anthropic/claude-opus-5:1",
      "fixtureId": "sup-007",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.039472,
      "latencyMs": 12981,
      "toolCalls": 7,
      "inputTokens": 8,
      "outputTokens": 769,
      "reasoningTokens": 59,
      "cacheReadTokens": 8464,
      "cacheWriteTokens": 2320
    },
    {
      "trialId": "scored:sup-007:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "sup-007",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000384318,
      "latencyMs": 8242,
      "toolCalls": 9,
      "inputTokens": 1641,
      "outputTokens": 994,
      "reasoningTokens": 0,
      "cacheReadTokens": 7680,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-007:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "rec-007",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01041184,
      "latencyMs": 29670,
      "toolCalls": 7,
      "inputTokens": 30,
      "outputTokens": 787,
      "reasoningTokens": 176,
      "cacheReadTokens": 6802,
      "cacheWriteTokens": 1367
    },
    {
      "trialId": "scored:sup-007:openai/gpt-5.6-sol:1",
      "fixtureId": "sup-007",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0094745,
      "latencyMs": 8594,
      "toolCalls": 6,
      "inputTokens": 795,
      "outputTokens": 308,
      "reasoningTokens": 68,
      "cacheReadTokens": 2685,
      "cacheWriteTokens": 1435
    },
    {
      "trialId": "scored:sup-007:alibaba/qwen3.8-flash:1",
      "fixtureId": "sup-007",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000954384,
      "latencyMs": 13744,
      "toolCalls": 8,
      "inputTokens": 24,
      "outputTokens": 762,
      "reasoningTokens": 318,
      "cacheReadTokens": 5959,
      "cacheWriteTokens": 1738
    },
    {
      "trialId": "scored:sup-007:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "sup-007",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.007691860000000001,
      "latencyMs": 16206,
      "toolCalls": 7,
      "inputTokens": 3402,
      "outputTokens": 614,
      "reasoningTokens": 159,
      "cacheReadTokens": 3185,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-007:anthropic/claude-sonnet-5:1",
      "fixtureId": "sup-007",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0170815,
      "latencyMs": 12796,
      "toolCalls": 6,
      "inputTokens": 6,
      "outputTokens": 739,
      "reasoningTokens": 86,
      "cacheReadTokens": 4785,
      "cacheWriteTokens": 3145
    },
    {
      "trialId": "scored:rec-001:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "rec-001",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000126646,
      "latencyMs": 4239,
      "toolCalls": 3,
      "inputTokens": 389,
      "outputTokens": 330,
      "reasoningTokens": 0,
      "cacheReadTokens": 3328,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-007:google/gemini-3.8-flash:1",
      "fixtureId": "sup-007",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0139455,
      "latencyMs": 16109,
      "toolCalls": 7,
      "inputTokens": 8654,
      "outputTokens": 255,
      "reasoningTokens": 1733,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-001:openai/gpt-5.6-sol:1",
      "fixtureId": "rec-001",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008916,
      "latencyMs": 7644,
      "toolCalls": 3,
      "inputTokens": 3273,
      "outputTokens": 208,
      "reasoningTokens": 29,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-001:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "rec-001",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.003873364,
      "latencyMs": 7801,
      "toolCalls": 4,
      "inputTokens": 1505,
      "outputTokens": 373,
      "reasoningTokens": 66,
      "cacheReadTokens": 3371,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-001:anthropic/claude-sonnet-5:1",
      "fixtureId": "rec-001",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.009704299999999999,
      "latencyMs": 8527,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 352,
      "reasoningTokens": 0,
      "cacheReadTokens": 3624,
      "cacheWriteTokens": 2179
    },
    {
      "trialId": "scored:rec-001:google/gemini-3.8-flash:1",
      "fixtureId": "rec-001",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0071445,
      "latencyMs": 10751,
      "toolCalls": 4,
      "inputTokens": 4531,
      "outputTokens": 274,
      "reasoningTokens": 725,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-001:alibaba/qwen3.8-flash:1",
      "fixtureId": "rec-001",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000497188,
      "latencyMs": 11082,
      "toolCalls": 3,
      "inputTokens": 18,
      "outputTokens": 454,
      "reasoningTokens": 120,
      "cacheReadTokens": 3033,
      "cacheWriteTokens": 880
    },
    {
      "trialId": "scored:rec-001:anthropic/claude-opus-5:1",
      "fixtureId": "rec-001",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.02358,
      "latencyMs": 9082,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 488,
      "reasoningTokens": 26,
      "cacheReadTokens": 4350,
      "cacheWriteTokens": 1364
    },
    {
      "trialId": "scored:sup-007:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "sup-007",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01452774,
      "latencyMs": 38478,
      "toolCalls": 10,
      "inputTokens": 30,
      "outputTokens": 902,
      "reasoningTokens": 450,
      "cacheReadTokens": 8122,
      "cacheWriteTokens": 1990
    },
    {
      "trialId": "scored:ext-004:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "ext-004",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000253814,
      "latencyMs": 6037,
      "toolCalls": 6,
      "inputTokens": 905,
      "outputTokens": 858,
      "reasoningTokens": 0,
      "cacheReadTokens": 3840,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-001:openai/gpt-5.6-luna:1",
      "fixtureId": "rec-001",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.001029,
      "latencyMs": 8081,
      "toolCalls": 3,
      "inputTokens": 3285,
      "outputTokens": 244,
      "reasoningTokens": 66,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-004:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "ext-004",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.007534032000000001,
      "latencyMs": 10484,
      "toolCalls": 6,
      "inputTokens": 3149,
      "outputTokens": 601,
      "reasoningTokens": 235,
      "cacheReadTokens": 1518,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-004:anthropic/claude-sonnet-5:1",
      "fixtureId": "ext-004",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0172949,
      "latencyMs": 11906,
      "toolCalls": 6,
      "inputTokens": 6,
      "outputTokens": 774,
      "reasoningTokens": 128,
      "cacheReadTokens": 4377,
      "cacheWriteTokens": 2955
    },
    {
      "trialId": "scored:rec-001:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "rec-001",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008410319999999999,
      "latencyMs": 25198,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 583,
      "reasoningTokens": 235,
      "cacheReadTokens": 4496,
      "cacheWriteTokens": 1076
    },
    {
      "trialId": "scored:ext-004:google/gemini-3.8-flash:1",
      "fixtureId": "ext-004",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.015956249999999998,
      "latencyMs": 21556,
      "toolCalls": 7,
      "inputTokens": 7065,
      "outputTokens": 368,
      "reasoningTokens": 2474,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-004:openai/gpt-5.6-luna:1",
      "fixtureId": "ext-004",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00103859,
      "latencyMs": 7800,
      "toolCalls": 6,
      "inputTokens": 745,
      "outputTokens": 375,
      "reasoningTokens": 66,
      "cacheReadTokens": 2282,
      "cacheWriteTokens": 1259
    },
    {
      "trialId": "scored:ext-004:openai/gpt-5.6-sol:1",
      "fixtureId": "ext-004",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.006771000000000001,
      "latencyMs": 10240,
      "toolCalls": 6,
      "inputTokens": 745,
      "outputTokens": 378,
      "reasoningTokens": 31,
      "cacheReadTokens": 3305,
      "cacheWriteTokens": 212
    },
    {
      "trialId": "scored:ext-004:alibaba/qwen3.8-flash:1",
      "fixtureId": "ext-004",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0008279660000000001,
      "latencyMs": 12403,
      "toolCalls": 6,
      "inputTokens": 18,
      "outputTokens": 681,
      "reasoningTokens": 340,
      "cacheReadTokens": 3426,
      "cacheWriteTokens": 1452
    },
    {
      "trialId": "scored:ext-004:anthropic/claude-opus-5:1",
      "fixtureId": "ext-004",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.036382750000000005,
      "latencyMs": 13005,
      "toolCalls": 6,
      "inputTokens": 6,
      "outputTokens": 775,
      "reasoningTokens": 79,
      "cacheReadTokens": 5093,
      "cacheWriteTokens": 1993
    },
    {
      "trialId": "scored:sup-006:openai/gpt-5.6-sol:1",
      "fixtureId": "sup-006",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.005882700000000001,
      "latencyMs": 10084,
      "toolCalls": 5,
      "inputTokens": 836,
      "outputTokens": 273,
      "reasoningTokens": 35,
      "cacheReadTokens": 3741,
      "cacheWriteTokens": 153
    },
    {
      "trialId": "scored:sup-006:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "sup-006",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.006598416,
      "latencyMs": 10323,
      "toolCalls": 5,
      "inputTokens": 3230,
      "outputTokens": 426,
      "reasoningTokens": 145,
      "cacheReadTokens": 1674,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-006:anthropic/claude-opus-5:1",
      "fixtureId": "sup-006",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.032805,
      "latencyMs": 10438,
      "toolCalls": 5,
      "inputTokens": 6,
      "outputTokens": 652,
      "reasoningTokens": 31,
      "cacheReadTokens": 5425,
      "cacheWriteTokens": 2078
    },
    {
      "trialId": "scored:sup-006:anthropic/claude-sonnet-5:1",
      "fixtureId": "sup-006",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0169757,
      "latencyMs": 11977,
      "toolCalls": 5,
      "inputTokens": 6,
      "outputTokens": 767,
      "reasoningTokens": 81,
      "cacheReadTokens": 4706,
      "cacheWriteTokens": 3017
    },
    {
      "trialId": "scored:sup-006:alibaba/qwen3.8-flash:1",
      "fixtureId": "sup-006",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000823232,
      "latencyMs": 14339,
      "toolCalls": 7,
      "inputTokens": 30,
      "outputTokens": 505,
      "reasoningTokens": 295,
      "cacheReadTokens": 7252,
      "cacheWriteTokens": 1632
    },
    {
      "trialId": "scored:ext-004:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "ext-004",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01230353,
      "latencyMs": 28929,
      "toolCalls": 7,
      "inputTokens": 18,
      "outputTokens": 815,
      "reasoningTokens": 371,
      "cacheReadTokens": 3509,
      "cacheWriteTokens": 1822
    },
    {
      "trialId": "scored:sup-006:google/gemini-3.8-flash:1",
      "fixtureId": "sup-006",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.012452250000000002,
      "latencyMs": 18553,
      "toolCalls": 7,
      "inputTokens": 9483,
      "outputTokens": 333,
      "reasoningTokens": 1091,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-006:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "sup-006",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000307057,
      "latencyMs": 7932,
      "toolCalls": 6,
      "inputTokens": 1335,
      "outputTokens": 805,
      "reasoningTokens": 0,
      "cacheReadTokens": 5888,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-010:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "ext-010",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000252149,
      "latencyMs": 7909,
      "toolCalls": 5,
      "inputTokens": 731,
      "outputTokens": 793,
      "reasoningTokens": 0,
      "cacheReadTokens": 5376,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-006:openai/gpt-5.6-luna:1",
      "fixtureId": "sup-006",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0010454300000000002,
      "latencyMs": 8338,
      "toolCalls": 5,
      "inputTokens": 836,
      "outputTokens": 300,
      "reasoningTokens": 100,
      "cacheReadTokens": 2574,
      "cacheWriteTokens": 1387
    },
    {
      "trialId": "scored:ext-010:openai/gpt-5.6-sol:1",
      "fixtureId": "ext-010",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.009362899999999999,
      "latencyMs": 8843,
      "toolCalls": 3,
      "inputTokens": 1545,
      "outputTokens": 259,
      "reasoningTokens": 73,
      "cacheReadTokens": 1077,
      "cacheWriteTokens": 1095
    },
    {
      "trialId": "scored:ext-010:google/gemini-3.8-flash:1",
      "fixtureId": "ext-010",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.014928750000000001,
      "latencyMs": 21012,
      "toolCalls": 5,
      "inputTokens": 8195,
      "outputTokens": 254,
      "reasoningTokens": 2088,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-010:anthropic/claude-sonnet-5:1",
      "fixtureId": "ext-010",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0158084,
      "latencyMs": 13938,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 594,
      "reasoningTokens": 255,
      "cacheReadTokens": 3807,
      "cacheWriteTokens": 2618
    },
    {
      "trialId": "scored:ext-010:openai/gpt-5.6-luna:1",
      "fixtureId": "ext-010",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0013744,
      "latencyMs": 8702,
      "toolCalls": 3,
      "inputTokens": 1545,
      "outputTokens": 269,
      "reasoningTokens": 148,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 2260
    },
    {
      "trialId": "scored:sup-006:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "sup-006",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01615697,
      "latencyMs": 43419,
      "toolCalls": 8,
      "inputTokens": 36,
      "outputTokens": 937,
      "reasoningTokens": 538,
      "cacheReadTokens": 9441,
      "cacheWriteTokens": 2252
    },
    {
      "trialId": "scored:ext-010:anthropic/claude-opus-5:1",
      "fixtureId": "ext-010",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.036665249999999996,
      "latencyMs": 16642,
      "toolCalls": 5,
      "inputTokens": 8,
      "outputTokens": 793,
      "reasoningTokens": 70,
      "cacheReadTokens": 7063,
      "cacheWriteTokens": 1843
    },
    {
      "trialId": "scored:ext-010:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "ext-010",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008071448,
      "latencyMs": 17153,
      "toolCalls": 3,
      "inputTokens": 3113,
      "outputTokens": 423,
      "reasoningTokens": 563,
      "cacheReadTokens": 1312,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-010:alibaba/qwen3.8-flash:1",
      "fixtureId": "ext-010",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000930216,
      "latencyMs": 23593,
      "toolCalls": 4,
      "inputTokens": 24,
      "outputTokens": 615,
      "reasoningTokens": 553,
      "cacheReadTokens": 5076,
      "cacheWriteTokens": 1481
    },
    {
      "trialId": "scored:sup-005:openai/gpt-5.6-sol:1",
      "fixtureId": "sup-005",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0151642,
      "latencyMs": 15489,
      "toolCalls": 6,
      "inputTokens": 962,
      "outputTokens": 353,
      "reasoningTokens": 100,
      "cacheReadTokens": 2826,
      "cacheWriteTokens": 3258
    },
    {
      "trialId": "scored:sup-005:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "sup-005",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000461638,
      "latencyMs": 16164,
      "toolCalls": 9,
      "inputTokens": 1301,
      "outputTokens": 1434,
      "reasoningTokens": 0,
      "cacheReadTokens": 10240,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-005:anthropic/claude-opus-5:1",
      "fixtureId": "sup-005",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0450665,
      "latencyMs": 20377,
      "toolCalls": 6,
      "inputTokens": 8,
      "outputTokens": 827,
      "reasoningTokens": 151,
      "cacheReadTokens": 8953,
      "cacheWriteTokens": 2576
    },
    {
      "trialId": "scored:sup-005:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "sup-005",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0114895,
      "latencyMs": 20553,
      "toolCalls": 6,
      "inputTokens": 4362,
      "outputTokens": 573,
      "reasoningTokens": 853,
      "cacheReadTokens": 1925,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-005:google/gemini-3.8-flash:1",
      "fixtureId": "sup-005",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.016371,
      "latencyMs": 21269,
      "toolCalls": 6,
      "inputTokens": 8923,
      "outputTokens": 459,
      "reasoningTokens": 2122,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-005:alibaba/qwen3.8-flash:1",
      "fixtureId": "sup-005",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0012496579999999998,
      "latencyMs": 22485,
      "toolCalls": 6,
      "inputTokens": 24,
      "outputTokens": 622,
      "reasoningTokens": 889,
      "cacheReadTokens": 6553,
      "cacheWriteTokens": 2154
    },
    {
      "trialId": "scored:ext-001:anthropic/claude-opus-5:1",
      "fixtureId": "ext-001",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0272865,
      "latencyMs": 8716,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 497,
      "reasoningTokens": 0,
      "cacheReadTokens": 5338,
      "cacheWriteTokens": 1946
    },
    {
      "trialId": "scored:ext-001:deepseek/deepseek-v4-pro-0813:1",
      "fixtureId": "ext-001",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.005046888,
      "latencyMs": 9093,
      "toolCalls": 2,
      "inputTokens": 1970,
      "outputTokens": 440,
      "reasoningTokens": 146,
      "cacheReadTokens": 2862,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-010:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "ext-010",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.011649650000000001,
      "latencyMs": 33106,
      "toolCalls": 4,
      "inputTokens": 24,
      "outputTokens": 763,
      "reasoningTokens": 459,
      "cacheReadTokens": 4895,
      "cacheWriteTokens": 1375
    },
    {
      "trialId": "scored:sup-005:openai/gpt-5.6-luna:1",
      "fixtureId": "sup-005",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00129504,
      "latencyMs": 12246,
      "toolCalls": 6,
      "inputTokens": 962,
      "outputTokens": 361,
      "reasoningTokens": 141,
      "cacheReadTokens": 4437,
      "cacheWriteTokens": 1646
    },
    {
      "trialId": "scored:sup-005:anthropic/claude-sonnet-5:1",
      "fixtureId": "sup-005",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0253093,
      "latencyMs": 19571,
      "toolCalls": 6,
      "inputTokens": 8,
      "outputTokens": 785,
      "reasoningTokens": 603,
      "cacheReadTokens": 8654,
      "cacheWriteTokens": 3873
    },
    {
      "trialId": "scored:ext-001:openai/gpt-5.6-luna:1",
      "fixtureId": "ext-001",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0009945499999999999,
      "latencyMs": 8985,
      "toolCalls": 2,
      "inputTokens": 883,
      "outputTokens": 279,
      "reasoningTokens": 79,
      "cacheReadTokens": 2455,
      "cacheWriteTokens": 1357
    },
    {
      "trialId": "scored:ext-001:anthropic/claude-sonnet-5:1",
      "fixtureId": "ext-001",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.013203699999999999,
      "latencyMs": 10530,
      "toolCalls": 2,
      "inputTokens": 6,
      "outputTokens": 537,
      "reasoningTokens": 0,
      "cacheReadTokens": 4521,
      "cacheWriteTokens": 2767
    },
    {
      "trialId": "scored:ext-001:google/gemini-3.8-flash:1",
      "fixtureId": "ext-001",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.009771,
      "latencyMs": 14342,
      "toolCalls": 2,
      "inputTokens": 4888,
      "outputTokens": 243,
      "reasoningTokens": 1385,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-001:alibaba/qwen3.8-flash:1",
      "fixtureId": "ext-001",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.001019342,
      "latencyMs": 15127,
      "toolCalls": 4,
      "inputTokens": 24,
      "outputTokens": 721,
      "reasoningTokens": 480,
      "cacheReadTokens": 5752,
      "cacheWriteTokens": 1795
    },
    {
      "trialId": "scored:ext-001:deepseek/deepseek-v4-flash-0731:1",
      "fixtureId": "ext-001",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00021798500000000002,
      "latencyMs": 4552,
      "toolCalls": 2,
      "inputTokens": 607,
      "outputTokens": 725,
      "reasoningTokens": 0,
      "cacheReadTokens": 4352,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-001:openai/gpt-5.6-sol:1",
      "fixtureId": "ext-001",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0088444,
      "latencyMs": 8939,
      "toolCalls": 2,
      "inputTokens": 883,
      "outputTokens": 308,
      "reasoningTokens": 22,
      "cacheReadTokens": 2417,
      "cacheWriteTokens": 1318
    },
    {
      "trialId": "scored:ext-003:alibaba/qwen3.8-flash:2",
      "fixtureId": "ext-003",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000489422,
      "latencyMs": 7145,
      "toolCalls": 2,
      "inputTokens": 18,
      "outputTokens": 321,
      "reasoningTokens": 112,
      "cacheReadTokens": 3602,
      "cacheWriteTokens": 1127
    },
    {
      "trialId": "scored:ext-003:openai/gpt-5.6-luna:2",
      "fixtureId": "ext-003",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00065208,
      "latencyMs": 7737,
      "toolCalls": 2,
      "inputTokens": 951,
      "outputTokens": 233,
      "reasoningTokens": 61,
      "cacheReadTokens": 3379,
      "cacheWriteTokens": 166
    },
    {
      "trialId": "scored:ext-003:anthropic/claude-opus-5:2",
      "fixtureId": "ext-003",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.024982250000000004,
      "latencyMs": 7761,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 409,
      "reasoningTokens": 0,
      "cacheReadTokens": 5517,
      "cacheWriteTokens": 1915
    },
    {
      "trialId": "scored:ext-003:openai/gpt-5.6-sol:2",
      "fixtureId": "ext-003",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0080938,
      "latencyMs": 7774,
      "toolCalls": 2,
      "inputTokens": 951,
      "outputTokens": 236,
      "reasoningTokens": 28,
      "cacheReadTokens": 2284,
      "cacheWriteTokens": 1238
    },
    {
      "trialId": "scored:ext-003:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "ext-003",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00797308,
      "latencyMs": 19664,
      "toolCalls": 4,
      "inputTokens": 24,
      "outputTokens": 515,
      "reasoningTokens": 129,
      "cacheReadTokens": 5374,
      "cacheWriteTokens": 1259
    },
    {
      "trialId": "scored:ext-003:google/gemini-3.8-flash:2",
      "fixtureId": "ext-003",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008439,
      "latencyMs": 11299,
      "toolCalls": 3,
      "inputTokens": 5382,
      "outputTokens": 172,
      "reasoningTokens": 1002,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-003:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "ext-003",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000176557,
      "latencyMs": 4818,
      "toolCalls": 3,
      "inputTokens": 537,
      "outputTokens": 489,
      "reasoningTokens": 0,
      "cacheReadTokens": 4352,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-003:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "ext-003",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.005775000000000001,
      "latencyMs": 6795,
      "toolCalls": 2,
      "inputTokens": 3154,
      "outputTokens": 329,
      "reasoningTokens": 61,
      "cacheReadTokens": 1530,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-005:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "sup-005",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.021597190000000002,
      "latencyMs": 55636,
      "toolCalls": 10,
      "inputTokens": 30,
      "outputTokens": 1143,
      "reasoningTokens": 956,
      "cacheReadTokens": 9607,
      "cacheWriteTokens": 2924
    },
    {
      "trialId": "scored:rec-006:openai/gpt-5.6-luna:2",
      "fixtureId": "rec-006",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00099619,
      "latencyMs": 8132,
      "toolCalls": 4,
      "inputTokens": 1556,
      "outputTokens": 262,
      "reasoningTokens": 74,
      "cacheReadTokens": 1027,
      "cacheWriteTokens": 1045
    },
    {
      "trialId": "scored:ext-003:anthropic/claude-sonnet-5:2",
      "fixtureId": "ext-003",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.013074200000000001,
      "latencyMs": 8946,
      "toolCalls": 2,
      "inputTokens": 6,
      "outputTokens": 508,
      "reasoningTokens": 9,
      "cacheReadTokens": 4711,
      "cacheWriteTokens": 2780
    },
    {
      "trialId": "scored:ext-001:alibaba/qwen3.8-max-0902:1",
      "fixtureId": "ext-001",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01270085,
      "latencyMs": 34704,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 882,
      "reasoningTokens": 377,
      "cacheReadTokens": 5655,
      "cacheWriteTokens": 1655
    },
    {
      "trialId": "scored:rec-006:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "rec-006",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00792022,
      "latencyMs": 16029,
      "toolCalls": 5,
      "inputTokens": 4228,
      "outputTokens": 462,
      "reasoningTokens": 117,
      "cacheReadTokens": 1055,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-006:anthropic/claude-sonnet-5:2",
      "fixtureId": "rec-006",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0149405,
      "latencyMs": 16437,
      "toolCalls": 6,
      "inputTokens": 8,
      "outputTokens": 605,
      "reasoningTokens": 100,
      "cacheReadTokens": 6310,
      "cacheWriteTokens": 2645
    },
    {
      "trialId": "scored:rec-006:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "rec-006",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000179203,
      "latencyMs": 5883,
      "toolCalls": 4,
      "inputTokens": 516,
      "outputTokens": 587,
      "reasoningTokens": 0,
      "cacheReadTokens": 3584,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-006:openai/gpt-5.6-sol:2",
      "fixtureId": "rec-006",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0091071,
      "latencyMs": 11085,
      "toolCalls": 4,
      "inputTokens": 1563,
      "outputTokens": 254,
      "reasoningTokens": 58,
      "cacheReadTokens": 1043,
      "cacheWriteTokens": 1061
    },
    {
      "trialId": "scored:rec-006:alibaba/qwen3.8-flash:2",
      "fixtureId": "rec-006",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0006981939999999999,
      "latencyMs": 11929,
      "toolCalls": 6,
      "inputTokens": 24,
      "outputTokens": 608,
      "reasoningTokens": 203,
      "cacheReadTokens": 4824,
      "cacheWriteTokens": 1180
    },
    {
      "trialId": "scored:rec-006:google/gemini-3.8-flash:2",
      "fixtureId": "rec-006",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0090615,
      "latencyMs": 12744,
      "toolCalls": 6,
      "inputTokens": 5332,
      "outputTokens": 385,
      "reasoningTokens": 965,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-007:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "ext-007",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00025317600000000003,
      "latencyMs": 6290,
      "toolCalls": 6,
      "inputTokens": 949,
      "outputTokens": 668,
      "reasoningTokens": 0,
      "cacheReadTokens": 5632,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-006:anthropic/claude-opus-5:2",
      "fixtureId": "rec-006",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.025756,
      "latencyMs": 9652,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 519,
      "reasoningTokens": 43,
      "cacheReadTokens": 4602,
      "cacheWriteTokens": 1500
    },
    {
      "trialId": "scored:ext-007:alibaba/qwen3.8-flash:2",
      "fixtureId": "ext-007",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0007790579999999999,
      "latencyMs": 12978,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 570,
      "reasoningTokens": 309,
      "cacheReadTokens": 5243,
      "cacheWriteTokens": 1391
    },
    {
      "trialId": "scored:rec-006:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "rec-006",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008410540000000001,
      "latencyMs": 23688,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 589,
      "reasoningTokens": 204,
      "cacheReadTokens": 4762,
      "cacheWriteTokens": 1118
    },
    {
      "trialId": "scored:ext-007:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "ext-007",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.006379956,
      "latencyMs": 9695,
      "toolCalls": 5,
      "inputTokens": 3105,
      "outputTokens": 440,
      "reasoningTokens": 104,
      "cacheReadTokens": 2889,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-007:anthropic/claude-opus-5:2",
      "fixtureId": "ext-007",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.03499225,
      "latencyMs": 14029,
      "toolCalls": 6,
      "inputTokens": 8,
      "outputTokens": 702,
      "reasoningTokens": 21,
      "cacheReadTokens": 7942,
      "cacheWriteTokens": 2065
    },
    {
      "trialId": "scored:ext-007:openai/gpt-5.6-sol:2",
      "fixtureId": "ext-007",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008302400000000001,
      "latencyMs": 10758,
      "toolCalls": 4,
      "inputTokens": 837,
      "outputTokens": 282,
      "reasoningTokens": 32,
      "cacheReadTokens": 2242,
      "cacheWriteTokens": 1216
    },
    {
      "trialId": "scored:ext-007:anthropic/claude-sonnet-5:2",
      "fixtureId": "ext-007",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0142788,
      "latencyMs": 11272,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 574,
      "reasoningTokens": 71,
      "cacheReadTokens": 4459,
      "cacheWriteTokens": 2770
    },
    {
      "trialId": "scored:sup-003:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "sup-003",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": false,
      "reasons": [
        "params.decision: substantive value or membership differs"
      ],
      "cost": 0.000503823,
      "latencyMs": 8448,
      "toolCalls": 4,
      "inputTokens": 2268,
      "outputTokens": 1815,
      "reasoningTokens": 0,
      "cacheReadTokens": 3840,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-007:google/gemini-3.8-flash:2",
      "fixtureId": "ext-007",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008903999999999999,
      "latencyMs": 13345,
      "toolCalls": 5,
      "inputTokens": 5337,
      "outputTokens": 229,
      "reasoningTokens": 1078,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-007:openai/gpt-5.6-luna:2",
      "fixtureId": "ext-007",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0009636,
      "latencyMs": 9926,
      "toolCalls": 4,
      "inputTokens": 837,
      "outputTokens": 281,
      "reasoningTokens": 86,
      "cacheReadTokens": 2265,
      "cacheWriteTokens": 1242
    },
    {
      "trialId": "scored:ext-007:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "ext-007",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01208142,
      "latencyMs": 33046,
      "toolCalls": 7,
      "inputTokens": 30,
      "outputTokens": 874,
      "reasoningTokens": 283,
      "cacheReadTokens": 6776,
      "cacheWriteTokens": 1571
    },
    {
      "trialId": "scored:sup-003:alibaba/qwen3.8-flash:2",
      "fixtureId": "sup-003",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00104563,
      "latencyMs": 22505,
      "toolCalls": 4,
      "inputTokens": 18,
      "outputTokens": 433,
      "reasoningTokens": 824,
      "cacheReadTokens": 3610,
      "cacheWriteTokens": 1971
    },
    {
      "trialId": "scored:sup-003:anthropic/claude-opus-5:2",
      "fixtureId": "sup-003",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.03897725,
      "latencyMs": 13675,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 868,
      "reasoningTokens": 24,
      "cacheReadTokens": 5307,
      "cacheWriteTokens": 2239
    },
    {
      "trialId": "scored:sup-003:google/gemini-3.8-flash:2",
      "fixtureId": "sup-003",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.011508750000000002,
      "latencyMs": 14311,
      "toolCalls": 5,
      "inputTokens": 6985,
      "outputTokens": 258,
      "reasoningTokens": 1414,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-003:anthropic/claude-sonnet-5:2",
      "fixtureId": "sup-003",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": false,
      "reasons": [
        "params.decision: substantive value or membership differs"
      ],
      "cost": 0.0184623,
      "latencyMs": 14311,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 560,
      "reasoningTokens": 375,
      "cacheReadTokens": 4589,
      "cacheWriteTokens": 3273
    },
    {
      "trialId": "scored:sup-003:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "sup-003",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.006632296,
      "latencyMs": 14679,
      "toolCalls": 4,
      "inputTokens": 3240,
      "outputTokens": 394,
      "reasoningTokens": 183,
      "cacheReadTokens": 1604,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-003:openai/gpt-5.6-sol:2",
      "fixtureId": "sup-003",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0061057,
      "latencyMs": 9151,
      "toolCalls": 4,
      "inputTokens": 846,
      "outputTokens": 284,
      "reasoningTokens": 41,
      "cacheReadTokens": 3581,
      "cacheWriteTokens": 179
    },
    {
      "trialId": "scored:rec-010:openai/gpt-5.6-sol:2",
      "fixtureId": "rec-010",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.009816199999999999,
      "latencyMs": 9146,
      "toolCalls": 6,
      "inputTokens": 1662,
      "outputTokens": 297,
      "reasoningTokens": 41,
      "cacheReadTokens": 1136,
      "cacheWriteTokens": 1154
    },
    {
      "trialId": "scored:sup-003:openai/gpt-5.6-luna:2",
      "fixtureId": "sup-003",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00099709,
      "latencyMs": 9501,
      "toolCalls": 4,
      "inputTokens": 846,
      "outputTokens": 275,
      "reasoningTokens": 94,
      "cacheReadTokens": 2467,
      "cacheWriteTokens": 1343
    },
    {
      "trialId": "scored:rec-010:openai/gpt-5.6-luna:2",
      "fixtureId": "rec-010",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0010518,
      "latencyMs": 9489,
      "toolCalls": 6,
      "inputTokens": 1662,
      "outputTokens": 269,
      "reasoningTokens": 68,
      "cacheReadTokens": 1150,
      "cacheWriteTokens": 1168
    },
    {
      "trialId": "scored:rec-010:anthropic/claude-opus-5:2",
      "fixtureId": "rec-010",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0368235,
      "latencyMs": 12448,
      "toolCalls": 7,
      "inputTokens": 8,
      "outputTokens": 805,
      "reasoningTokens": 41,
      "cacheReadTokens": 7367,
      "cacheWriteTokens": 1912
    },
    {
      "trialId": "scored:rec-010:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "rec-010",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000209735,
      "latencyMs": 4931,
      "toolCalls": 6,
      "inputTokens": 805,
      "outputTokens": 643,
      "reasoningTokens": 0,
      "cacheReadTokens": 3584,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-003:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "sup-003",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01278653,
      "latencyMs": 36883,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 658,
      "reasoningTokens": 523,
      "cacheReadTokens": 5809,
      "cacheWriteTokens": 1866
    },
    {
      "trialId": "scored:rec-010:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "rec-010",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.006876716000000001,
      "latencyMs": 24001,
      "toolCalls": 8,
      "inputTokens": 2981,
      "outputTokens": 587,
      "reasoningTokens": 106,
      "cacheReadTokens": 4489,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-010:google/gemini-3.8-flash:2",
      "fixtureId": "rec-010",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01025625,
      "latencyMs": 14624,
      "toolCalls": 8,
      "inputTokens": 7765,
      "outputTokens": 269,
      "reasoningTokens": 913,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-010:alibaba/qwen3.8-flash:2",
      "fixtureId": "rec-010",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000889166,
      "latencyMs": 15193,
      "toolCalls": 9,
      "inputTokens": 30,
      "outputTokens": 688,
      "reasoningTokens": 361,
      "cacheReadTokens": 6921,
      "cacheWriteTokens": 1403
    },
    {
      "trialId": "scored:rec-010:anthropic/claude-sonnet-5:2",
      "fixtureId": "rec-010",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0164359,
      "latencyMs": 17279,
      "toolCalls": 6,
      "inputTokens": 6,
      "outputTokens": 753,
      "reasoningTokens": 112,
      "cacheReadTokens": 4157,
      "cacheWriteTokens": 2777
    },
    {
      "trialId": "scored:ext-005:openai/gpt-5.6-luna:2",
      "fixtureId": "ext-005",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0012152,
      "latencyMs": 6872,
      "toolCalls": 4,
      "inputTokens": 1560,
      "outputTokens": 245,
      "reasoningTokens": 76,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 2072
    },
    {
      "trialId": "scored:ext-005:anthropic/claude-sonnet-5:2",
      "fixtureId": "ext-005",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0137022,
      "latencyMs": 9385,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 595,
      "reasoningTokens": 83,
      "cacheReadTokens": 3876,
      "cacheWriteTokens": 2454
    },
    {
      "trialId": "scored:ext-005:google/gemini-3.8-flash:2",
      "fixtureId": "ext-005",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0083745,
      "latencyMs": 11838,
      "toolCalls": 5,
      "inputTokens": 5261,
      "outputTokens": 206,
      "reasoningTokens": 975,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-005:openai/gpt-5.6-sol:2",
      "fixtureId": "ext-005",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.010613,
      "latencyMs": 11061,
      "toolCalls": 4,
      "inputTokens": 2574,
      "outputTokens": 252,
      "reasoningTokens": 36,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 1034
    },
    {
      "trialId": "scored:ext-005:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "ext-005",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.005534056,
      "latencyMs": 14338,
      "toolCalls": 4,
      "inputTokens": 2711,
      "outputTokens": 316,
      "reasoningTokens": 163,
      "cacheReadTokens": 1334,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-005:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "ext-005",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00022477899999999998,
      "latencyMs": 6545,
      "toolCalls": 5,
      "inputTokens": 544,
      "outputTokens": 707,
      "reasoningTokens": 0,
      "cacheReadTokens": 5376,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-005:anthropic/claude-opus-5:2",
      "fixtureId": "ext-005",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.029925000000000004,
      "latencyMs": 15593,
      "toolCalls": 6,
      "inputTokens": 8,
      "outputTokens": 597,
      "reasoningTokens": 36,
      "cacheReadTokens": 7045,
      "cacheWriteTokens": 1686
    },
    {
      "trialId": "scored:ext-005:alibaba/qwen3.8-flash:2",
      "fixtureId": "ext-005",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000568512,
      "latencyMs": 13175,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 439,
      "reasoningTokens": 161,
      "cacheReadTokens": 4717,
      "cacheWriteTokens": 1036
    },
    {
      "trialId": "scored:rec-010:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "rec-010",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0131164,
      "latencyMs": 44209,
      "toolCalls": 11,
      "inputTokens": 30,
      "outputTokens": 892,
      "reasoningTokens": 368,
      "cacheReadTokens": 7420,
      "cacheWriteTokens": 1694
    },
    {
      "trialId": "scored:ext-005:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "ext-005",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.011206009999999999,
      "latencyMs": 34661,
      "toolCalls": 7,
      "inputTokens": 30,
      "outputTokens": 741,
      "reasoningTokens": 349,
      "cacheReadTokens": 6403,
      "cacheWriteTokens": 1407
    },
    {
      "trialId": "scored:rec-009:anthropic/claude-opus-5:2",
      "fixtureId": "rec-009",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.03723475,
      "latencyMs": 11530,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 720,
      "reasoningTokens": 136,
      "cacheReadTokens": 5422,
      "cacheWriteTokens": 2095
    },
    {
      "trialId": "scored:rec-009:anthropic/claude-sonnet-5:2",
      "fixtureId": "rec-009",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": false,
      "reasons": [
        "params.exceptionCodes: substantive value or membership differs"
      ],
      "cost": 0.0240557,
      "latencyMs": 18564,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 863,
      "reasoningTokens": 581,
      "cacheReadTokens": 4706,
      "cacheWriteTokens": 3465
    },
    {
      "trialId": "scored:rec-009:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "rec-009",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008760004,
      "latencyMs": 18552,
      "toolCalls": 5,
      "inputTokens": 3504,
      "outputTokens": 619,
      "reasoningTokens": 392,
      "cacheReadTokens": 2981,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-009:alibaba/qwen3.8-flash:2",
      "fixtureId": "rec-009",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.001085162,
      "latencyMs": 18565,
      "toolCalls": 6,
      "inputTokens": 24,
      "outputTokens": 810,
      "reasoningTokens": 605,
      "cacheReadTokens": 5642,
      "cacheWriteTokens": 1630
    },
    {
      "trialId": "scored:rec-009:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "rec-009",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00030114,
      "latencyMs": 7389,
      "toolCalls": 5,
      "inputTokens": 684,
      "outputTokens": 996,
      "reasoningTokens": 0,
      "cacheReadTokens": 6912,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-009:openai/gpt-5.6-sol:2",
      "fixtureId": "rec-009",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0124536,
      "latencyMs": 12853,
      "toolCalls": 4,
      "inputTokens": 885,
      "outputTokens": 345,
      "reasoningTokens": 57,
      "cacheReadTokens": 1093,
      "cacheWriteTokens": 2578
    },
    {
      "trialId": "scored:rec-009:openai/gpt-5.6-luna:2",
      "fixtureId": "rec-009",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0011919,
      "latencyMs": 9734,
      "toolCalls": 4,
      "inputTokens": 885,
      "outputTokens": 337,
      "reasoningTokens": 178,
      "cacheReadTokens": 2470,
      "cacheWriteTokens": 1390
    },
    {
      "trialId": "scored:sup-008:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "sup-008",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000321269,
      "latencyMs": 8341,
      "toolCalls": 7,
      "inputTokens": 933,
      "outputTokens": 1121,
      "reasoningTokens": 0,
      "cacheReadTokens": 5632,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-009:google/gemini-3.8-flash:2",
      "fixtureId": "rec-009",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.020925,
      "latencyMs": 26543,
      "toolCalls": 5,
      "inputTokens": 10075,
      "outputTokens": 441,
      "reasoningTokens": 3124,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-008:anthropic/claude-sonnet-5:2",
      "fixtureId": "sup-008",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0288187,
      "latencyMs": 17333,
      "toolCalls": 7,
      "inputTokens": 6,
      "outputTokens": 1304,
      "reasoningTokens": 341,
      "cacheReadTokens": 5896,
      "cacheWriteTokens": 4471
    },
    {
      "trialId": "scored:sup-008:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "sup-008",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.011256740000000001,
      "latencyMs": 19462,
      "toolCalls": 8,
      "inputTokens": 4556,
      "outputTokens": 785,
      "reasoningTokens": 496,
      "cacheReadTokens": 3865,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-008:openai/gpt-5.6-sol:2",
      "fixtureId": "sup-008",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.007054500000000001,
      "latencyMs": 8528,
      "toolCalls": 7,
      "inputTokens": 9,
      "outputTokens": 318,
      "reasoningTokens": 95,
      "cacheReadTokens": 3720,
      "cacheWriteTokens": 865
    },
    {
      "trialId": "scored:sup-008:openai/gpt-5.6-luna:2",
      "fixtureId": "sup-008",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00111955,
      "latencyMs": 9367,
      "toolCalls": 7,
      "inputTokens": 12,
      "outputTokens": 517,
      "reasoningTokens": 124,
      "cacheReadTokens": 5685,
      "cacheWriteTokens": 937
    },
    {
      "trialId": "scored:sup-008:google/gemini-3.8-flash:2",
      "fixtureId": "sup-008",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01739475,
      "latencyMs": 35347,
      "toolCalls": 7,
      "inputTokens": 7338,
      "outputTokens": 606,
      "reasoningTokens": 2565,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-008:anthropic/claude-opus-5:2",
      "fixtureId": "sup-008",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": false,
      "reasons": [
        "submission.params.decision: required property missing"
      ],
      "cost": 0.06478175,
      "latencyMs": 20517,
      "toolCalls": 7,
      "inputTokens": 6,
      "outputTokens": 1444,
      "reasoningTokens": 132,
      "cacheReadTokens": 6616,
      "cacheWriteTokens": 3527
    },
    {
      "trialId": "scored:sup-008:alibaba/qwen3.8-flash:2",
      "fixtureId": "sup-008",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0015935320000000002,
      "latencyMs": 22427,
      "toolCalls": 7,
      "inputTokens": 18,
      "outputTokens": 1434,
      "reasoningTokens": 520,
      "cacheReadTokens": 4392,
      "cacheWriteTokens": 3010
    },
    {
      "trialId": "scored:rec-009:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "rec-009",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01527193,
      "latencyMs": 49445,
      "toolCalls": 6,
      "inputTokens": 24,
      "outputTokens": 864,
      "reasoningTokens": 649,
      "cacheReadTokens": 5829,
      "cacheWriteTokens": 2062
    },
    {
      "trialId": "scored:rec-003:anthropic/claude-opus-5:2",
      "fixtureId": "rec-003",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.03178175,
      "latencyMs": 9966,
      "toolCalls": 5,
      "inputTokens": 6,
      "outputTokens": 654,
      "reasoningTokens": 34,
      "cacheReadTokens": 5066,
      "cacheWriteTokens": 1923
    },
    {
      "trialId": "scored:rec-003:openai/gpt-5.6-luna:2",
      "fixtureId": "rec-003",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0008287500000000001,
      "latencyMs": 9750,
      "toolCalls": 5,
      "inputTokens": 753,
      "outputTokens": 359,
      "reasoningTokens": 99,
      "cacheReadTokens": 3390,
      "cacheWriteTokens": 243
    },
    {
      "trialId": "scored:rec-003:openai/gpt-5.6-sol:2",
      "fixtureId": "rec-003",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.009266,
      "latencyMs": 11031,
      "toolCalls": 5,
      "inputTokens": 753,
      "outputTokens": 377,
      "reasoningTokens": 37,
      "cacheReadTokens": 2300,
      "cacheWriteTokens": 1264
    },
    {
      "trialId": "scored:rec-003:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "rec-003",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.006615576,
      "latencyMs": 9575,
      "toolCalls": 6,
      "inputTokens": 3092,
      "outputTokens": 531,
      "reasoningTokens": 77,
      "cacheReadTokens": 2874,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-003:anthropic/claude-sonnet-5:2",
      "fixtureId": "rec-003",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0166345,
      "latencyMs": 12750,
      "toolCalls": 5,
      "inputTokens": 6,
      "outputTokens": 761,
      "reasoningTokens": 91,
      "cacheReadTokens": 4350,
      "cacheWriteTokens": 2893
    },
    {
      "trialId": "scored:rec-003:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "rec-003",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000256124,
      "latencyMs": 5366,
      "toolCalls": 7,
      "inputTokens": 691,
      "outputTokens": 792,
      "reasoningTokens": 0,
      "cacheReadTokens": 5888,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-008:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "sup-008",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01962743,
      "latencyMs": 52448,
      "toolCalls": 7,
      "inputTokens": 18,
      "outputTokens": 1389,
      "reasoningTokens": 527,
      "cacheReadTokens": 4429,
      "cacheWriteTokens": 2937
    },
    {
      "trialId": "scored:ext-002:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "ext-002",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000198751,
      "latencyMs": 4807,
      "toolCalls": 3,
      "inputTokens": 542,
      "outputTokens": 655,
      "reasoningTokens": 0,
      "cacheReadTokens": 4096,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-003:alibaba/qwen3.8-flash:2",
      "fixtureId": "rec-003",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000822552,
      "latencyMs": 14926,
      "toolCalls": 7,
      "inputTokens": 24,
      "outputTokens": 723,
      "reasoningTokens": 241,
      "cacheReadTokens": 5152,
      "cacheWriteTokens": 1416
    },
    {
      "trialId": "scored:rec-003:google/gemini-3.8-flash:2",
      "fixtureId": "rec-003",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.011445,
      "latencyMs": 15615,
      "toolCalls": 7,
      "inputTokens": 7235,
      "outputTokens": 558,
      "reasoningTokens": 1047,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-002:openai/gpt-5.6-luna:2",
      "fixtureId": "ext-002",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0010673500000000001,
      "latencyMs": 8204,
      "toolCalls": 3,
      "inputTokens": 785,
      "outputTokens": 339,
      "reasoningTokens": 78,
      "cacheReadTokens": 2110,
      "cacheWriteTokens": 1471
    },
    {
      "trialId": "scored:ext-002:alibaba/qwen3.8-flash:2",
      "fixtureId": "ext-002",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000596414,
      "latencyMs": 9554,
      "toolCalls": 3,
      "inputTokens": 18,
      "outputTokens": 394,
      "reasoningTokens": 215,
      "cacheReadTokens": 3444,
      "cacheWriteTokens": 1261
    },
    {
      "trialId": "scored:ext-002:google/gemini-3.8-flash:2",
      "fixtureId": "ext-002",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0099255,
      "latencyMs": 21562,
      "toolCalls": 4,
      "inputTokens": 5729,
      "outputTokens": 425,
      "reasoningTokens": 1076,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-002:openai/gpt-5.6-sol:2",
      "fixtureId": "ext-002",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008783599999999999,
      "latencyMs": 9100,
      "toolCalls": 3,
      "inputTokens": 785,
      "outputTokens": 334,
      "reasoningTokens": 28,
      "cacheReadTokens": 2293,
      "cacheWriteTokens": 1254
    },
    {
      "trialId": "scored:ext-002:anthropic/claude-opus-5:2",
      "fixtureId": "ext-002",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.026729000000000003,
      "latencyMs": 10181,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 532,
      "reasoningTokens": 0,
      "cacheReadTokens": 4973,
      "cacheWriteTokens": 1746
    },
    {
      "trialId": "scored:ext-002:anthropic/claude-sonnet-5:2",
      "fixtureId": "ext-002",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0143032,
      "latencyMs": 10917,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 632,
      "reasoningTokens": 36,
      "cacheReadTokens": 4256,
      "cacheWriteTokens": 2704
    },
    {
      "trialId": "scored:ext-002:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "ext-002",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.006588911999999999,
      "latencyMs": 12406,
      "toolCalls": 3,
      "inputTokens": 3073,
      "outputTokens": 454,
      "reasoningTokens": 169,
      "cacheReadTokens": 1488,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-003:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "rec-003",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01167775,
      "latencyMs": 35361,
      "toolCalls": 9,
      "inputTokens": 30,
      "outputTokens": 822,
      "reasoningTokens": 231,
      "cacheReadTokens": 7175,
      "cacheWriteTokens": 1632
    },
    {
      "trialId": "scored:sup-004:anthropic/claude-opus-5:2",
      "fixtureId": "sup-004",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0475835,
      "latencyMs": 15699,
      "toolCalls": 5,
      "inputTokens": 6,
      "outputTokens": 1062,
      "reasoningTokens": 91,
      "cacheReadTokens": 5457,
      "cacheWriteTokens": 2560
    },
    {
      "trialId": "scored:sup-004:openai/gpt-5.6-luna:2",
      "fixtureId": "sup-004",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0012262999999999998,
      "latencyMs": 9021,
      "toolCalls": 5,
      "inputTokens": 844,
      "outputTokens": 354,
      "reasoningTokens": 159,
      "cacheReadTokens": 2770,
      "cacheWriteTokens": 1546
    },
    {
      "trialId": "scored:sup-004:anthropic/claude-sonnet-5:2",
      "fixtureId": "sup-004",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0165921,
      "latencyMs": 10630,
      "toolCalls": 5,
      "inputTokens": 6,
      "outputTokens": 639,
      "reasoningTokens": 128,
      "cacheReadTokens": 4738,
      "cacheWriteTokens": 3185
    },
    {
      "trialId": "scored:sup-004:google/gemini-3.8-flash:2",
      "fixtureId": "sup-004",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00986025,
      "latencyMs": 11699,
      "toolCalls": 5,
      "inputTokens": 5027,
      "outputTokens": 301,
      "reasoningTokens": 1323,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-004:alibaba/qwen3.8-flash:2",
      "fixtureId": "sup-004",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0009198160000000001,
      "latencyMs": 16457,
      "toolCalls": 6,
      "inputTokens": 24,
      "outputTokens": 669,
      "reasoningTokens": 371,
      "cacheReadTokens": 5786,
      "cacheWriteTokens": 1673
    },
    {
      "trialId": "scored:sup-004:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "sup-004",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00030451299999999996,
      "latencyMs": 7464,
      "toolCalls": 6,
      "inputTokens": 1022,
      "outputTokens": 897,
      "reasoningTokens": 0,
      "cacheReadTokens": 6400,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-004:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "sup-004",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.007926599999999999,
      "latencyMs": 12552,
      "toolCalls": 7,
      "inputTokens": 3554,
      "outputTokens": 580,
      "reasoningTokens": 179,
      "cacheReadTokens": 5220,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-002:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "rec-002",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00021750399999999998,
      "latencyMs": 5295,
      "toolCalls": 5,
      "inputTokens": 911,
      "outputTokens": 524,
      "reasoningTokens": 0,
      "cacheReadTokens": 4864,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-004:openai/gpt-5.6-sol:2",
      "fixtureId": "sup-004",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.009975000000000001,
      "latencyMs": 10222,
      "toolCalls": 5,
      "inputTokens": 844,
      "outputTokens": 343,
      "reasoningTokens": 66,
      "cacheReadTokens": 2685,
      "cacheWriteTokens": 1464
    },
    {
      "trialId": "scored:ext-002:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "ext-002",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.011801689999999998,
      "latencyMs": 35951,
      "toolCalls": 5,
      "inputTokens": 30,
      "outputTokens": 750,
      "reasoningTokens": 365,
      "cacheReadTokens": 6907,
      "cacheWriteTokens": 1551
    },
    {
      "trialId": "scored:rec-002:anthropic/claude-opus-5:2",
      "fixtureId": "rec-002",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.024480000000000002,
      "latencyMs": 8450,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 477,
      "reasoningTokens": 19,
      "cacheReadTokens": 4700,
      "cacheWriteTokens": 1552
    },
    {
      "trialId": "scored:rec-002:anthropic/claude-sonnet-5:2",
      "fixtureId": "rec-002",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0120318,
      "latencyMs": 9753,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 455,
      "reasoningTokens": 45,
      "cacheReadTokens": 3974,
      "cacheWriteTokens": 2490
    },
    {
      "trialId": "scored:rec-002:openai/gpt-5.6-sol:2",
      "fixtureId": "rec-002",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008844,
      "latencyMs": 11661,
      "toolCalls": 4,
      "inputTokens": 1602,
      "outputTokens": 246,
      "reasoningTokens": 30,
      "cacheReadTokens": 1050,
      "cacheWriteTokens": 1068
    },
    {
      "trialId": "scored:sup-004:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "sup-004",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01221703,
      "latencyMs": 34829,
      "toolCalls": 7,
      "inputTokens": 24,
      "outputTokens": 784,
      "reasoningTokens": 278,
      "cacheReadTokens": 5909,
      "cacheWriteTokens": 1917
    },
    {
      "trialId": "scored:rec-002:openai/gpt-5.6-luna:2",
      "fixtureId": "rec-002",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00103665,
      "latencyMs": 9457,
      "toolCalls": 4,
      "inputTokens": 1605,
      "outputTokens": 270,
      "reasoningTokens": 83,
      "cacheReadTokens": 1065,
      "cacheWriteTokens": 1083
    },
    {
      "trialId": "scored:rec-002:alibaba/qwen3.8-flash:2",
      "fixtureId": "rec-002",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000586056,
      "latencyMs": 10845,
      "toolCalls": 4,
      "inputTokens": 18,
      "outputTokens": 610,
      "reasoningTokens": 102,
      "cacheReadTokens": 3246,
      "cacheWriteTokens": 983
    },
    {
      "trialId": "scored:rec-002:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "rec-002",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.007122720000000001,
      "latencyMs": 10834,
      "toolCalls": 4,
      "inputTokens": 4055,
      "outputTokens": 379,
      "reasoningTokens": 68,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-002:google/gemini-3.8-flash:2",
      "fixtureId": "rec-002",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.009169499999999999,
      "latencyMs": 12043,
      "toolCalls": 5,
      "inputTokens": 5146,
      "outputTokens": 346,
      "reasoningTokens": 1070,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-002:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "rec-002",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00877532,
      "latencyMs": 29379,
      "toolCalls": 6,
      "inputTokens": 24,
      "outputTokens": 635,
      "reasoningTokens": 173,
      "cacheReadTokens": 4996,
      "cacheWriteTokens": 1212
    },
    {
      "trialId": "scored:sup-001:openai/gpt-5.6-luna:2",
      "fixtureId": "sup-001",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0009883399999999999,
      "latencyMs": 10213,
      "toolCalls": 5,
      "inputTokens": 750,
      "outputTokens": 286,
      "reasoningTokens": 93,
      "cacheReadTokens": 2452,
      "cacheWriteTokens": 1338
    },
    {
      "trialId": "scored:sup-001:openai/gpt-5.6-sol:2",
      "fixtureId": "sup-001",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008947,
      "latencyMs": 12209,
      "toolCalls": 5,
      "inputTokens": 750,
      "outputTokens": 315,
      "reasoningTokens": 50,
      "cacheReadTokens": 2435,
      "cacheWriteTokens": 1324
    },
    {
      "trialId": "scored:sup-001:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "sup-001",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000262374,
      "latencyMs": 7437,
      "toolCalls": 6,
      "inputTokens": 714,
      "outputTokens": 798,
      "reasoningTokens": 0,
      "cacheReadTokens": 6144,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-001:anthropic/claude-opus-5:2",
      "fixtureId": "sup-001",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.03288225,
      "latencyMs": 11075,
      "toolCalls": 5,
      "inputTokens": 6,
      "outputTokens": 627,
      "reasoningTokens": 66,
      "cacheReadTokens": 5117,
      "cacheWriteTokens": 2075
    },
    {
      "trialId": "scored:sup-001:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "sup-001",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.007536276,
      "latencyMs": 12963,
      "toolCalls": 6,
      "inputTokens": 3307,
      "outputTokens": 530,
      "reasoningTokens": 238,
      "cacheReadTokens": 2949,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-001:google/gemini-3.8-flash:2",
      "fixtureId": "sup-001",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01221,
      "latencyMs": 18413,
      "toolCalls": 5,
      "inputTokens": 8660,
      "outputTokens": 416,
      "reasoningTokens": 1108,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-001:anthropic/claude-sonnet-5:2",
      "fixtureId": "sup-001",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0147691,
      "latencyMs": 12336,
      "toolCalls": 5,
      "inputTokens": 6,
      "outputTokens": 554,
      "reasoningTokens": 100,
      "cacheReadTokens": 4398,
      "cacheWriteTokens": 2935
    },
    {
      "trialId": "scored:sup-001:alibaba/qwen3.8-flash:2",
      "fixtureId": "sup-001",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0009792639999999999,
      "latencyMs": 17424,
      "toolCalls": 6,
      "inputTokens": 24,
      "outputTokens": 801,
      "reasoningTokens": 375,
      "cacheReadTokens": 5494,
      "cacheWriteTokens": 1674
    },
    {
      "trialId": "scored:sup-009:openai/gpt-5.6-sol:2",
      "fixtureId": "sup-009",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.011397899999999999,
      "latencyMs": 14655,
      "toolCalls": 6,
      "inputTokens": 892,
      "outputTokens": 392,
      "reasoningTokens": 89,
      "cacheReadTokens": 3057,
      "cacheWriteTokens": 1677
    },
    {
      "trialId": "scored:sup-001:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "sup-001",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.010435920000000001,
      "latencyMs": 29443,
      "toolCalls": 7,
      "inputTokens": 24,
      "outputTokens": 697,
      "reasoningTokens": 219,
      "cacheReadTokens": 5526,
      "cacheWriteTokens": 1581
    },
    {
      "trialId": "scored:sup-009:anthropic/claude-opus-5:2",
      "fixtureId": "sup-009",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.04753725,
      "latencyMs": 17570,
      "toolCalls": 6,
      "inputTokens": 6,
      "outputTokens": 1012,
      "reasoningTokens": 84,
      "cacheReadTokens": 5902,
      "cacheWriteTokens": 2745
    },
    {
      "trialId": "scored:sup-009:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "sup-009",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008170624,
      "latencyMs": 11210,
      "toolCalls": 6,
      "inputTokens": 3720,
      "outputTokens": 548,
      "reasoningTokens": 254,
      "cacheReadTokens": 1916,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-009:openai/gpt-5.6-luna:2",
      "fixtureId": "sup-009",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00123323,
      "latencyMs": 11234,
      "toolCalls": 6,
      "inputTokens": 892,
      "outputTokens": 371,
      "reasoningTokens": 114,
      "cacheReadTokens": 3029,
      "cacheWriteTokens": 1649
    },
    {
      "trialId": "scored:sup-009:google/gemini-3.8-flash:2",
      "fixtureId": "sup-009",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01403475,
      "latencyMs": 17942,
      "toolCalls": 7,
      "inputTokens": 8268,
      "outputTokens": 459,
      "reasoningTokens": 1630,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-009:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "sup-009",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000377182,
      "latencyMs": 7177,
      "toolCalls": 7,
      "inputTokens": 1868,
      "outputTokens": 1022,
      "reasoningTokens": 0,
      "cacheReadTokens": 5632,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-005:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "sup-005",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00033940500000000003,
      "latencyMs": 10082,
      "toolCalls": 6,
      "inputTokens": 761,
      "outputTokens": 1161,
      "reasoningTokens": 0,
      "cacheReadTokens": 7424,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-005:openai/gpt-5.6-luna:2",
      "fixtureId": "sup-005",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0009621499999999998,
      "latencyMs": 11552,
      "toolCalls": 6,
      "inputTokens": 962,
      "outputTokens": 340,
      "reasoningTokens": 141,
      "cacheReadTokens": 5815,
      "cacheWriteTokens": 305
    },
    {
      "trialId": "scored:sup-009:anthropic/claude-sonnet-5:2",
      "fixtureId": "sup-009",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0216386,
      "latencyMs": 15348,
      "toolCalls": 6,
      "inputTokens": 6,
      "outputTokens": 1001,
      "reasoningTokens": 151,
      "cacheReadTokens": 5183,
      "cacheWriteTokens": 3628
    },
    {
      "trialId": "scored:sup-009:alibaba/qwen3.8-flash:2",
      "fixtureId": "sup-009",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.001075222,
      "latencyMs": 16102,
      "toolCalls": 7,
      "inputTokens": 24,
      "outputTokens": 837,
      "reasoningTokens": 336,
      "cacheReadTokens": 6342,
      "cacheWriteTokens": 2093
    },
    {
      "trialId": "scored:sup-005:anthropic/claude-sonnet-5:2",
      "fixtureId": "sup-005",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0222385,
      "latencyMs": 18407,
      "toolCalls": 6,
      "inputTokens": 8,
      "outputTokens": 768,
      "reasoningTokens": 375,
      "cacheReadTokens": 8400,
      "cacheWriteTokens": 3645
    },
    {
      "trialId": "scored:sup-005:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "sup-005",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.014248432000000002,
      "latencyMs": 20725,
      "toolCalls": 7,
      "inputTokens": 5983,
      "outputTokens": 651,
      "reasoningTokens": 899,
      "cacheReadTokens": 4838,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-005:anthropic/claude-opus-5:2",
      "fixtureId": "sup-005",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.04132025,
      "latencyMs": 16648,
      "toolCalls": 6,
      "inputTokens": 8,
      "outputTokens": 639,
      "reasoningTokens": 196,
      "cacheReadTokens": 9023,
      "cacheWriteTokens": 2543
    },
    {
      "trialId": "scored:sup-009:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "sup-009",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.016059280000000002,
      "latencyMs": 48730,
      "toolCalls": 9,
      "inputTokens": 30,
      "outputTokens": 992,
      "reasoningTokens": 454,
      "cacheReadTokens": 8284,
      "cacheWriteTokens": 2366
    },
    {
      "trialId": "scored:sup-005:alibaba/qwen3.8-flash:2",
      "fixtureId": "sup-005",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.001236478,
      "latencyMs": 18161,
      "toolCalls": 7,
      "inputTokens": 24,
      "outputTokens": 801,
      "reasoningTokens": 596,
      "cacheReadTokens": 6553,
      "cacheWriteTokens": 2356
    },
    {
      "trialId": "scored:sup-005:openai/gpt-5.6-sol:2",
      "fixtureId": "sup-005",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0112006,
      "latencyMs": 19404,
      "toolCalls": 6,
      "inputTokens": 962,
      "outputTokens": 366,
      "reasoningTokens": 72,
      "cacheReadTokens": 4383,
      "cacheWriteTokens": 1608
    },
    {
      "trialId": "scored:ext-009:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "ext-009",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000167452,
      "latencyMs": 5618,
      "toolCalls": 3,
      "inputTokens": 632,
      "outputTokens": 476,
      "reasoningTokens": 0,
      "cacheReadTokens": 3328,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-009:anthropic/claude-sonnet-5:2",
      "fixtureId": "ext-009",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0108105,
      "latencyMs": 9731,
      "toolCalls": 2,
      "inputTokens": 6,
      "outputTokens": 460,
      "reasoningTokens": 0,
      "cacheReadTokens": 3605,
      "cacheWriteTokens": 2191
    },
    {
      "trialId": "scored:ext-009:alibaba/qwen3.8-flash:2",
      "fixtureId": "ext-009",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00046674600000000003,
      "latencyMs": 9895,
      "toolCalls": 2,
      "inputTokens": 18,
      "outputTokens": 348,
      "reasoningTokens": 147,
      "cacheReadTokens": 3076,
      "cacheWriteTokens": 910
    },
    {
      "trialId": "scored:ext-009:openai/gpt-5.6-luna:2",
      "fixtureId": "ext-009",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0009864000000000001,
      "latencyMs": 10178,
      "toolCalls": 2,
      "inputTokens": 3354,
      "outputTokens": 244,
      "reasoningTokens": 19,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-009:google/gemini-3.8-flash:2",
      "fixtureId": "ext-009",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00682275,
      "latencyMs": 12118,
      "toolCalls": 3,
      "inputTokens": 4472,
      "outputTokens": 200,
      "reasoningTokens": 725,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-005:google/gemini-3.8-flash:2",
      "fixtureId": "sup-005",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.025728749999999998,
      "latencyMs": 38573,
      "toolCalls": 10,
      "inputTokens": 19235,
      "outputTokens": 608,
      "reasoningTokens": 2406,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-009:openai/gpt-5.6-sol:2",
      "fixtureId": "ext-009",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00632,
      "latencyMs": 4670,
      "toolCalls": 2,
      "inputTokens": 2385,
      "outputTokens": 145,
      "reasoningTokens": 10,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-009:anthropic/claude-opus-5:2",
      "fixtureId": "ext-009",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0218535,
      "latencyMs": 8577,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 442,
      "reasoningTokens": 0,
      "cacheReadTokens": 4422,
      "cacheWriteTokens": 1370
    },
    {
      "trialId": "scored:ext-006:openai/gpt-5.6-sol:2",
      "fixtureId": "ext-006",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008484100000000001,
      "latencyMs": 7955,
      "toolCalls": 4,
      "inputTokens": 751,
      "outputTokens": 305,
      "reasoningTokens": 38,
      "cacheReadTokens": 2273,
      "cacheWriteTokens": 1239
    },
    {
      "trialId": "scored:ext-009:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "ext-009",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.005356428000000001,
      "latencyMs": 12165,
      "toolCalls": 3,
      "inputTokens": 2618,
      "outputTokens": 374,
      "reasoningTokens": 92,
      "cacheReadTokens": 1257,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-006:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "ext-006",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00023008399999999997,
      "latencyMs": 4447,
      "toolCalls": 6,
      "inputTokens": 935,
      "outputTokens": 688,
      "reasoningTokens": 0,
      "cacheReadTokens": 3840,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-006:anthropic/claude-sonnet-5:2",
      "fixtureId": "ext-006",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0150151,
      "latencyMs": 12783,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 610,
      "reasoningTokens": 105,
      "cacheReadTokens": 4303,
      "cacheWriteTokens": 2797
    },
    {
      "trialId": "scored:sup-005:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "sup-005",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.02163899,
      "latencyMs": 71075,
      "toolCalls": 9,
      "inputTokens": 30,
      "outputTokens": 986,
      "reasoningTokens": 1158,
      "cacheReadTokens": 10147,
      "cacheWriteTokens": 2796
    },
    {
      "trialId": "scored:ext-006:anthropic/claude-opus-5:2",
      "fixtureId": "ext-006",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.03528075,
      "latencyMs": 15649,
      "toolCalls": 6,
      "inputTokens": 8,
      "outputTokens": 714,
      "reasoningTokens": 37,
      "cacheReadTokens": 7744,
      "cacheWriteTokens": 2015
    },
    {
      "trialId": "scored:ext-006:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "ext-006",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.006184948,
      "latencyMs": 9580,
      "toolCalls": 5,
      "inputTokens": 3043,
      "outputTokens": 438,
      "reasoningTokens": 78,
      "cacheReadTokens": 2837,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-006:alibaba/qwen3.8-flash:2",
      "fixtureId": "ext-006",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0007172020000000001,
      "latencyMs": 10643,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 591,
      "reasoningTokens": 180,
      "cacheReadTokens": 5212,
      "cacheWriteTokens": 1338
    },
    {
      "trialId": "scored:ext-009:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "ext-009",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.007923280000000001,
      "latencyMs": 24272,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 497,
      "reasoningTokens": 219,
      "cacheReadTokens": 4584,
      "cacheWriteTokens": 1120
    },
    {
      "trialId": "scored:ext-006:google/gemini-3.8-flash:2",
      "fixtureId": "ext-006",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008418,
      "latencyMs": 11564,
      "toolCalls": 5,
      "inputTokens": 5544,
      "outputTokens": 282,
      "reasoningTokens": 854,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-006:openai/gpt-5.6-sol:2",
      "fixtureId": "sup-006",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.011412299999999998,
      "latencyMs": 9483,
      "toolCalls": 5,
      "inputTokens": 836,
      "outputTokens": 270,
      "reasoningTokens": 37,
      "cacheReadTokens": 1339,
      "cacheWriteTokens": 2561
    },
    {
      "trialId": "scored:ext-006:openai/gpt-5.6-luna:2",
      "fixtureId": "ext-006",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00076172,
      "latencyMs": 11271,
      "toolCalls": 4,
      "inputTokens": 751,
      "outputTokens": 340,
      "reasoningTokens": 74,
      "cacheReadTokens": 3336,
      "cacheWriteTokens": 192
    },
    {
      "trialId": "scored:sup-006:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "sup-006",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00033651,
      "latencyMs": 6886,
      "toolCalls": 7,
      "inputTokens": 1701,
      "outputTokens": 722,
      "reasoningTokens": 0,
      "cacheReadTokens": 6912,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-006:anthropic/claude-sonnet-5:2",
      "fixtureId": "sup-006",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.016288700000000003,
      "latencyMs": 13642,
      "toolCalls": 5,
      "inputTokens": 6,
      "outputTokens": 740,
      "reasoningTokens": 48,
      "cacheReadTokens": 4696,
      "cacheWriteTokens": 2983
    },
    {
      "trialId": "scored:sup-006:openai/gpt-5.6-luna:2",
      "fixtureId": "sup-006",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00102516,
      "latencyMs": 8289,
      "toolCalls": 5,
      "inputTokens": 836,
      "outputTokens": 274,
      "reasoningTokens": 107,
      "cacheReadTokens": 2588,
      "cacheWriteTokens": 1396
    },
    {
      "trialId": "scored:sup-006:google/gemini-3.8-flash:2",
      "fixtureId": "sup-006",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.011476499999999999,
      "latencyMs": 14896,
      "toolCalls": 7,
      "inputTokens": 8047,
      "outputTokens": 342,
      "reasoningTokens": 1109,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-006:anthropic/claude-opus-5:2",
      "fixtureId": "sup-006",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.04402975,
      "latencyMs": 15196,
      "toolCalls": 6,
      "inputTokens": 8,
      "outputTokens": 956,
      "reasoningTokens": 50,
      "cacheReadTokens": 8392,
      "cacheWriteTokens": 2343
    },
    {
      "trialId": "scored:sup-006:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "sup-006",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0073321160000000005,
      "latencyMs": 18592,
      "toolCalls": 6,
      "inputTokens": 3321,
      "outputTokens": 568,
      "reasoningTokens": 142,
      "cacheReadTokens": 3109,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-006:alibaba/qwen3.8-flash:2",
      "fixtureId": "sup-006",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0011124479999999998,
      "latencyMs": 19423,
      "toolCalls": 7,
      "inputTokens": 48,
      "outputTokens": 703,
      "reasoningTokens": 409,
      "cacheReadTokens": 12383,
      "cacheWriteTokens": 1920
    },
    {
      "trialId": "scored:ext-006:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "ext-006",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01354748,
      "latencyMs": 39765,
      "toolCalls": 7,
      "inputTokens": 24,
      "outputTokens": 869,
      "reasoningTokens": 510,
      "cacheReadTokens": 5444,
      "cacheWriteTokens": 1720
    },
    {
      "trialId": "scored:rec-005:openai/gpt-5.6-luna:2",
      "fixtureId": "rec-005",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000939,
      "latencyMs": 8782,
      "toolCalls": 3,
      "inputTokens": 3075,
      "outputTokens": 204,
      "reasoningTokens": 66,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-005:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "rec-005",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000191685,
      "latencyMs": 5174,
      "toolCalls": 4,
      "inputTokens": 728,
      "outputTokens": 493,
      "reasoningTokens": 0,
      "cacheReadTokens": 4352,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-005:openai/gpt-5.6-sol:2",
      "fixtureId": "rec-005",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008308,
      "latencyMs": 8803,
      "toolCalls": 3,
      "inputTokens": 3064,
      "outputTokens": 185,
      "reasoningTokens": 33,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-005:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "rec-005",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00587532,
      "latencyMs": 8798,
      "toolCalls": 3,
      "inputTokens": 3500,
      "outputTokens": 273,
      "reasoningTokens": 44,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-005:anthropic/claude-sonnet-5:2",
      "fixtureId": "rec-005",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0104136,
      "latencyMs": 10628,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 374,
      "reasoningTokens": 60,
      "cacheReadTokens": 3508,
      "cacheWriteTokens": 2144
    },
    {
      "trialId": "scored:rec-001:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "rec-001",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00013305400000000002,
      "latencyMs": 4472,
      "toolCalls": 3,
      "inputTokens": 425,
      "outputTokens": 354,
      "reasoningTokens": 0,
      "cacheReadTokens": 3328,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-005:anthropic/claude-opus-5:2",
      "fixtureId": "rec-005",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.022296500000000004,
      "latencyMs": 9503,
      "toolCalls": 4,
      "inputTokens": 8,
      "outputTokens": 417,
      "reasoningTokens": 33,
      "cacheReadTokens": 6163,
      "cacheWriteTokens": 1268
    },
    {
      "trialId": "scored:rec-005:google/gemini-3.8-flash:2",
      "fixtureId": "rec-005",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.006072750000000001,
      "latencyMs": 9495,
      "toolCalls": 4,
      "inputTokens": 3937,
      "outputTokens": 248,
      "reasoningTokens": 584,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-005:alibaba/qwen3.8-flash:2",
      "fixtureId": "rec-005",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000456478,
      "latencyMs": 7971,
      "toolCalls": 3,
      "inputTokens": 18,
      "outputTokens": 405,
      "reasoningTokens": 100,
      "cacheReadTokens": 2928,
      "cacheWriteTokens": 847
    },
    {
      "trialId": "scored:sup-006:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "sup-006",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01180336,
      "latencyMs": 35534,
      "toolCalls": 7,
      "inputTokens": 30,
      "outputTokens": 651,
      "reasoningTokens": 375,
      "cacheReadTokens": 7558,
      "cacheWriteTokens": 1721
    },
    {
      "trialId": "scored:rec-001:anthropic/claude-sonnet-5:2",
      "fixtureId": "rec-001",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0102068,
      "latencyMs": 9133,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 395,
      "reasoningTokens": 0,
      "cacheReadTokens": 3624,
      "cacheWriteTokens": 2208
    },
    {
      "trialId": "scored:rec-005:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "rec-005",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00825417,
      "latencyMs": 27465,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 667,
      "reasoningTokens": 141,
      "cacheReadTokens": 4401,
      "cacheWriteTokens": 1044
    },
    {
      "trialId": "scored:rec-001:openai/gpt-5.6-luna:2",
      "fixtureId": "rec-001",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0009898,
      "latencyMs": 14332,
      "toolCalls": 3,
      "inputTokens": 3293,
      "outputTokens": 197,
      "reasoningTokens": 79,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-001:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "rec-001",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00875419,
      "latencyMs": 29527,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 528,
      "reasoningTokens": 336,
      "cacheReadTokens": 4557,
      "cacheWriteTokens": 1099
    },
    {
      "trialId": "scored:rec-001:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "rec-001",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0045294919999999995,
      "latencyMs": 9792,
      "toolCalls": 3,
      "inputTokens": 2413,
      "outputTokens": 295,
      "reasoningTokens": 31,
      "cacheReadTokens": 1213,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-001:alibaba/qwen3.8-flash:2",
      "fixtureId": "rec-001",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000508904,
      "latencyMs": 9816,
      "toolCalls": 3,
      "inputTokens": 18,
      "outputTokens": 428,
      "reasoningTokens": 164,
      "cacheReadTokens": 3024,
      "cacheWriteTokens": 897
    },
    {
      "trialId": "scored:rec-001:anthropic/claude-opus-5:2",
      "fixtureId": "rec-001",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0217295,
      "latencyMs": 9158,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 420,
      "reasoningTokens": 31,
      "cacheReadTokens": 4349,
      "cacheWriteTokens": 1320
    },
    {
      "trialId": "scored:sup-010:openai/gpt-5.6-sol:2",
      "fixtureId": "sup-010",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0093665,
      "latencyMs": 9567,
      "toolCalls": 4,
      "inputTokens": 904,
      "outputTokens": 314,
      "reasoningTokens": 31,
      "cacheReadTokens": 2655,
      "cacheWriteTokens": 1431
    },
    {
      "trialId": "scored:rec-001:openai/gpt-5.6-sol:2",
      "fixtureId": "rec-001",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.009087999999999999,
      "latencyMs": 9582,
      "toolCalls": 3,
      "inputTokens": 3299,
      "outputTokens": 209,
      "reasoningTokens": 40,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-010:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "sup-010",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000218054,
      "latencyMs": 4769,
      "toolCalls": 4,
      "inputTokens": 423,
      "outputTokens": 770,
      "reasoningTokens": 0,
      "cacheReadTokens": 4864,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-010:openai/gpt-5.6-luna:2",
      "fixtureId": "sup-010",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00110681,
      "latencyMs": 7905,
      "toolCalls": 4,
      "inputTokens": 904,
      "outputTokens": 308,
      "reasoningTokens": 110,
      "cacheReadTokens": 2708,
      "cacheWriteTokens": 1481
    },
    {
      "trialId": "scored:sup-010:alibaba/qwen3.8-flash:2",
      "fixtureId": "sup-010",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000804534,
      "latencyMs": 14501,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 659,
      "reasoningTokens": 178,
      "cacheReadTokens": 5294,
      "cacheWriteTokens": 1613
    },
    {
      "trialId": "scored:sup-010:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "sup-010",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.006603652,
      "latencyMs": 8294,
      "toolCalls": 4,
      "inputTokens": 3392,
      "outputTokens": 372,
      "reasoningTokens": 146,
      "cacheReadTokens": 1703,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-002:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "sup-002",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000274012,
      "latencyMs": 6636,
      "toolCalls": 6,
      "inputTokens": 1222,
      "outputTokens": 692,
      "reasoningTokens": 0,
      "cacheReadTokens": 5376,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-010:google/gemini-3.8-flash:2",
      "fixtureId": "sup-010",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.018549,
      "latencyMs": 22544,
      "toolCalls": 5,
      "inputTokens": 11127,
      "outputTokens": 362,
      "reasoningTokens": 2359,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-010:anthropic/claude-sonnet-5:2",
      "fixtureId": "sup-010",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0189615,
      "latencyMs": 14584,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 856,
      "reasoningTokens": 111,
      "cacheReadTokens": 4860,
      "cacheWriteTokens": 3323
    },
    {
      "trialId": "scored:sup-010:anthropic/claude-opus-5:2",
      "fixtureId": "sup-010",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.041243,
      "latencyMs": 12852,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 925,
      "reasoningTokens": 17,
      "cacheReadTokens": 5576,
      "cacheWriteTokens": 2380
    },
    {
      "trialId": "scored:rec-001:google/gemini-3.8-flash:2",
      "fixtureId": "rec-001",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00652575,
      "latencyMs": 56904,
      "toolCalls": 4,
      "inputTokens": 4071,
      "outputTokens": 210,
      "reasoningTokens": 716,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-002:anthropic/claude-opus-5:2",
      "fixtureId": "sup-002",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.03798,
      "latencyMs": 12617,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 803,
      "reasoningTokens": 90,
      "cacheReadTokens": 5125,
      "cacheWriteTokens": 2090
    },
    {
      "trialId": "scored:sup-010:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "sup-010",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01275322,
      "latencyMs": 38161,
      "toolCalls": 7,
      "inputTokens": 24,
      "outputTokens": 849,
      "reasoningTokens": 322,
      "cacheReadTokens": 5966,
      "cacheWriteTokens": 1866
    },
    {
      "trialId": "scored:sup-002:openai/gpt-5.6-sol:2",
      "fixtureId": "sup-002",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0079764,
      "latencyMs": 11710,
      "toolCalls": 4,
      "inputTokens": 803,
      "outputTokens": 246,
      "reasoningTokens": 34,
      "cacheReadTokens": 2302,
      "cacheWriteTokens": 1244
    },
    {
      "trialId": "scored:sup-002:alibaba/qwen3.8-flash:2",
      "fixtureId": "sup-002",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000658954,
      "latencyMs": 11411,
      "toolCalls": 4,
      "inputTokens": 18,
      "outputTokens": 591,
      "reasoningTokens": 124,
      "cacheReadTokens": 3514,
      "cacheWriteTokens": 1319
    },
    {
      "trialId": "scored:sup-002:anthropic/claude-sonnet-5:2",
      "fixtureId": "sup-002",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0145332,
      "latencyMs": 12614,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 588,
      "reasoningTokens": 64,
      "cacheReadTokens": 4406,
      "cacheWriteTokens": 2848
    },
    {
      "trialId": "scored:sup-002:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "sup-002",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00777744,
      "latencyMs": 9535,
      "toolCalls": 4,
      "inputTokens": 4560,
      "outputTokens": 354,
      "reasoningTokens": 90,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-002:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "sup-002",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00942217,
      "latencyMs": 24630,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 623,
      "reasoningTokens": 171,
      "cacheReadTokens": 5251,
      "cacheWriteTokens": 1487
    },
    {
      "trialId": "scored:sup-002:openai/gpt-5.6-luna:2",
      "fixtureId": "sup-002",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00093713,
      "latencyMs": 8684,
      "toolCalls": 4,
      "inputTokens": 803,
      "outputTokens": 273,
      "reasoningTokens": 70,
      "cacheReadTokens": 2334,
      "cacheWriteTokens": 1273
    },
    {
      "trialId": "scored:ext-001:openai/gpt-5.6-luna:2",
      "fixtureId": "ext-001",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00081503,
      "latencyMs": 9563,
      "toolCalls": 2,
      "inputTokens": 883,
      "outputTokens": 344,
      "reasoningTokens": 81,
      "cacheReadTokens": 3559,
      "cacheWriteTokens": 229
    },
    {
      "trialId": "scored:ext-001:google/gemini-3.8-flash:2",
      "fixtureId": "ext-001",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.009729,
      "latencyMs": 12912,
      "toolCalls": 3,
      "inputTokens": 5972,
      "outputTokens": 240,
      "reasoningTokens": 1160,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-001:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "ext-001",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00021538499999999999,
      "latencyMs": 5879,
      "toolCalls": 2,
      "inputTokens": 605,
      "outputTokens": 709,
      "reasoningTokens": 0,
      "cacheReadTokens": 4352,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-001:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "ext-001",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0063245160000000005,
      "latencyMs": 8042,
      "toolCalls": 2,
      "inputTokens": 3296,
      "outputTokens": 303,
      "reasoningTokens": 178,
      "cacheReadTokens": 1569,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-002:google/gemini-3.8-flash:2",
      "fixtureId": "sup-002",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.016036500000000002,
      "latencyMs": 25280,
      "toolCalls": 7,
      "inputTokens": 10752,
      "outputTokens": 374,
      "reasoningTokens": 1752,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-001:alibaba/qwen3.8-flash:2",
      "fixtureId": "ext-001",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0007854419999999999,
      "latencyMs": 11149,
      "toolCalls": 2,
      "inputTokens": 18,
      "outputTokens": 626,
      "reasoningTokens": 293,
      "cacheReadTokens": 3652,
      "cacheWriteTokens": 1461
    },
    {
      "trialId": "scored:ext-001:openai/gpt-5.6-sol:2",
      "fixtureId": "ext-001",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.011245,
      "latencyMs": 12024,
      "toolCalls": 2,
      "inputTokens": 883,
      "outputTokens": 298,
      "reasoningTokens": 17,
      "cacheReadTokens": 1295,
      "cacheWriteTokens": 2428
    },
    {
      "trialId": "scored:ext-010:openai/gpt-5.6-luna:2",
      "fixtureId": "ext-010",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0013825,
      "latencyMs": 8087,
      "toolCalls": 3,
      "inputTokens": 1547,
      "outputTokens": 272,
      "reasoningTokens": 146,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 2286
    },
    {
      "trialId": "scored:ext-001:anthropic/claude-sonnet-5:2",
      "fixtureId": "ext-001",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0123537,
      "latencyMs": 8278,
      "toolCalls": 2,
      "inputTokens": 6,
      "outputTokens": 452,
      "reasoningTokens": 0,
      "cacheReadTokens": 4521,
      "cacheWriteTokens": 2767
    },
    {
      "trialId": "scored:ext-001:anthropic/claude-opus-5:2",
      "fixtureId": "ext-001",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.031057750000000002,
      "latencyMs": 10182,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 635,
      "reasoningTokens": 9,
      "cacheReadTokens": 5343,
      "cacheWriteTokens": 1961
    },
    {
      "trialId": "scored:ext-010:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "ext-010",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000284759,
      "latencyMs": 6996,
      "toolCalls": 4,
      "inputTokens": 666,
      "outputTokens": 1015,
      "reasoningTokens": 0,
      "cacheReadTokens": 5632,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-010:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "ext-010",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.006334284000000001,
      "latencyMs": 14581,
      "toolCalls": 3,
      "inputTokens": 2805,
      "outputTokens": 395,
      "reasoningTokens": 255,
      "cacheReadTokens": 1311,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-010:anthropic/claude-sonnet-5:2",
      "fixtureId": "ext-010",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.013175700000000002,
      "latencyMs": 9968,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 442,
      "reasoningTokens": 166,
      "cacheReadTokens": 3806,
      "cacheWriteTokens": 2529
    },
    {
      "trialId": "scored:ext-010:anthropic/claude-opus-5:2",
      "fixtureId": "ext-010",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.034298499999999996,
      "latencyMs": 16182,
      "toolCalls": 4,
      "inputTokens": 8,
      "outputTokens": 658,
      "reasoningTokens": 152,
      "cacheReadTokens": 6817,
      "cacheWriteTokens": 1696
    },
    {
      "trialId": "scored:ext-001:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "ext-001",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01059804,
      "latencyMs": 30610,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 750,
      "reasoningTokens": 206,
      "cacheReadTokens": 5612,
      "cacheWriteTokens": 1544
    },
    {
      "trialId": "scored:ext-010:openai/gpt-5.6-sol:2",
      "fixtureId": "ext-010",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.009764199999999999,
      "latencyMs": 12474,
      "toolCalls": 3,
      "inputTokens": 1545,
      "outputTokens": 275,
      "reasoningTokens": 92,
      "cacheReadTokens": 1096,
      "cacheWriteTokens": 1114
    },
    {
      "trialId": "scored:rec-004:anthropic/claude-opus-5:2",
      "fixtureId": "rec-004",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.030396250000000003,
      "latencyMs": 11678,
      "toolCalls": 5,
      "inputTokens": 6,
      "outputTokens": 600,
      "reasoningTokens": 25,
      "cacheReadTokens": 5170,
      "cacheWriteTokens": 1945
    },
    {
      "trialId": "scored:ext-010:alibaba/qwen3.8-flash:2",
      "fixtureId": "ext-010",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0008538859999999999,
      "latencyMs": 16178,
      "toolCalls": 4,
      "inputTokens": 24,
      "outputTokens": 526,
      "reasoningTokens": 503,
      "cacheReadTokens": 4976,
      "cacheWriteTokens": 1434
    },
    {
      "trialId": "scored:ext-010:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "ext-010",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01259474,
      "latencyMs": 38890,
      "toolCalls": 4,
      "inputTokens": 24,
      "outputTokens": 752,
      "reasoningTokens": 542,
      "cacheReadTokens": 4972,
      "cacheWriteTokens": 1575
    },
    {
      "trialId": "scored:rec-004:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "rec-004",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00025812999999999997,
      "latencyMs": 5598,
      "toolCalls": 5,
      "inputTokens": 833,
      "outputTokens": 758,
      "reasoningTokens": 0,
      "cacheReadTokens": 5632,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-004:openai/gpt-5.6-luna:2",
      "fixtureId": "rec-004",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00098997,
      "latencyMs": 12557,
      "toolCalls": 4,
      "inputTokens": 782,
      "outputTokens": 322,
      "reasoningTokens": 75,
      "cacheReadTokens": 2271,
      "cacheWriteTokens": 1247
    },
    {
      "trialId": "scored:rec-004:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "rec-004",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008259240000000001,
      "latencyMs": 16396,
      "toolCalls": 4,
      "inputTokens": 4508,
      "outputTokens": 478,
      "reasoningTokens": 105,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-004:alibaba/qwen3.8-flash:2",
      "fixtureId": "rec-004",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0008522320000000001,
      "latencyMs": 17343,
      "toolCalls": 6,
      "inputTokens": 24,
      "outputTokens": 840,
      "reasoningTokens": 188,
      "cacheReadTokens": 5227,
      "cacheWriteTokens": 1408
    },
    {
      "trialId": "scored:rec-004:google/gemini-3.8-flash:2",
      "fixtureId": "rec-004",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0119145,
      "latencyMs": 17948,
      "toolCalls": 6,
      "inputTokens": 7741,
      "outputTokens": 486,
      "reasoningTokens": 1143,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-004:openai/gpt-5.6-sol:2",
      "fixtureId": "rec-004",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008880200000000001,
      "latencyMs": 9695,
      "toolCalls": 4,
      "inputTokens": 782,
      "outputTokens": 342,
      "reasoningTokens": 36,
      "cacheReadTokens": 2256,
      "cacheWriteTokens": 1234
    },
    {
      "trialId": "scored:rec-004:anthropic/claude-sonnet-5:2",
      "fixtureId": "rec-004",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0143569,
      "latencyMs": 10110,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 585,
      "reasoningTokens": 70,
      "cacheReadTokens": 4337,
      "cacheWriteTokens": 2771
    },
    {
      "trialId": "scored:rec-004:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "rec-004",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01099321,
      "latencyMs": 33398,
      "toolCalls": 6,
      "inputTokens": 24,
      "outputTokens": 885,
      "reasoningTokens": 198,
      "cacheReadTokens": 5263,
      "cacheWriteTokens": 1421
    },
    {
      "trialId": "scored:ext-010:google/gemini-3.8-flash:2",
      "fixtureId": "ext-010",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01206525,
      "latencyMs": 90179,
      "toolCalls": 4,
      "inputTokens": 6172,
      "outputTokens": 383,
      "reasoningTokens": 1600,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-008:openai/gpt-5.6-sol:2",
      "fixtureId": "rec-008",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.009680399999999999,
      "latencyMs": 9014,
      "toolCalls": 4,
      "inputTokens": 1665,
      "outputTokens": 298,
      "reasoningTokens": 35,
      "cacheReadTokens": 1102,
      "cacheWriteTokens": 1120
    },
    {
      "trialId": "scored:rec-008:anthropic/claude-sonnet-5:2",
      "fixtureId": "rec-008",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0143461,
      "latencyMs": 12109,
      "toolCalls": 4,
      "inputTokens": 6,
      "outputTokens": 589,
      "reasoningTokens": 100,
      "cacheReadTokens": 4108,
      "cacheWriteTokens": 2649
    },
    {
      "trialId": "scored:rec-008:google/gemini-3.8-flash:2",
      "fixtureId": "rec-008",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00799575,
      "latencyMs": 12314,
      "toolCalls": 5,
      "inputTokens": 5246,
      "outputTokens": 261,
      "reasoningTokens": 822,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-008:anthropic/claude-opus-5:2",
      "fixtureId": "rec-008",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0338475,
      "latencyMs": 12709,
      "toolCalls": 6,
      "inputTokens": 8,
      "outputTokens": 645,
      "reasoningTokens": 80,
      "cacheReadTokens": 7315,
      "cacheWriteTokens": 1924
    },
    {
      "trialId": "scored:rec-008:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "rec-008",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.005060571999999999,
      "latencyMs": 14027,
      "toolCalls": 7,
      "inputTokens": 1885,
      "outputTokens": 497,
      "reasoningTokens": 92,
      "cacheReadTokens": 5453,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-008:alibaba/qwen3.8-flash:2",
      "fixtureId": "rec-008",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0007751920000000001,
      "latencyMs": 14485,
      "toolCalls": 6,
      "inputTokens": 24,
      "outputTokens": 682,
      "reasoningTokens": 222,
      "cacheReadTokens": 4992,
      "cacheWriteTokens": 1333
    },
    {
      "trialId": "scored:rec-008:openai/gpt-5.6-luna:2",
      "fixtureId": "rec-008",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0010784599999999998,
      "latencyMs": 8990,
      "toolCalls": 4,
      "inputTokens": 1663,
      "outputTokens": 281,
      "reasoningTokens": 83,
      "cacheReadTokens": 1128,
      "cacheWriteTokens": 1146
    },
    {
      "trialId": "scored:rec-008:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "rec-008",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00030578,
      "latencyMs": 11410,
      "toolCalls": 7,
      "inputTokens": 1242,
      "outputTokens": 796,
      "reasoningTokens": 0,
      "cacheReadTokens": 6400,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-007:openai/gpt-5.6-luna:2",
      "fixtureId": "rec-007",
      "model": "openai/gpt-5.6-luna",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0011379699999999999,
      "latencyMs": 9343,
      "toolCalls": 5,
      "inputTokens": 1748,
      "outputTokens": 300,
      "reasoningTokens": 83,
      "cacheReadTokens": 1201,
      "cacheWriteTokens": 1219
    },
    {
      "trialId": "scored:rec-007:openai/gpt-5.6-sol:2",
      "fixtureId": "rec-007",
      "model": "openai/gpt-5.6-sol",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.010575600000000001,
      "latencyMs": 10850,
      "toolCalls": 5,
      "inputTokens": 1748,
      "outputTokens": 331,
      "reasoningTokens": 49,
      "cacheReadTokens": 1198,
      "cacheWriteTokens": 1216
    },
    {
      "trialId": "scored:rec-008:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "rec-008",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01044618,
      "latencyMs": 27541,
      "toolCalls": 7,
      "inputTokens": 24,
      "outputTokens": 731,
      "reasoningTokens": 262,
      "cacheReadTokens": 5104,
      "cacheWriteTokens": 1429
    },
    {
      "trialId": "scored:rec-007:alibaba/qwen3.8-flash:2",
      "fixtureId": "rec-007",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0008004919999999999,
      "latencyMs": 12760,
      "toolCalls": 7,
      "inputTokens": 24,
      "outputTokens": 661,
      "reasoningTokens": 245,
      "cacheReadTokens": 5227,
      "cacheWriteTokens": 1436
    },
    {
      "trialId": "scored:rec-007:anthropic/claude-sonnet-5:2",
      "fixtureId": "rec-007",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0139006,
      "latencyMs": 13038,
      "toolCalls": 5,
      "inputTokens": 6,
      "outputTokens": 592,
      "reasoningTokens": 38,
      "cacheReadTokens": 4218,
      "cacheWriteTokens": 2698
    },
    {
      "trialId": "scored:rec-007:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "rec-007",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000254039,
      "latencyMs": 7161,
      "toolCalls": 6,
      "inputTokens": 577,
      "outputTokens": 835,
      "reasoningTokens": 0,
      "cacheReadTokens": 5888,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-007:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "rec-007",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.010831128,
      "latencyMs": 21792,
      "toolCalls": 7,
      "inputTokens": 5953,
      "outputTokens": 550,
      "reasoningTokens": 181,
      "cacheReadTokens": 1782,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-008:anthropic/claude-opus-5:2",
      "fixtureId": "ext-008",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0193155,
      "latencyMs": 9102,
      "toolCalls": 3,
      "inputTokens": 6,
      "outputTokens": 331,
      "reasoningTokens": 36,
      "cacheReadTokens": 4346,
      "cacheWriteTokens": 1270
    },
    {
      "trialId": "scored:rec-007:google/gemini-3.8-flash:2",
      "fixtureId": "rec-007",
      "model": "google/gemini-3.8-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00886275,
      "latencyMs": 11380,
      "toolCalls": 6,
      "inputTokens": 5517,
      "outputTokens": 306,
      "reasoningTokens": 954,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-007:anthropic/claude-opus-5:2",
      "fixtureId": "rec-007",
      "model": "anthropic/claude-opus-5",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0325155,
      "latencyMs": 11598,
      "toolCalls": 6,
      "inputTokens": 8,
      "outputTokens": 645,
      "reasoningTokens": 40,
      "cacheReadTokens": 7351,
      "cacheWriteTokens": 1868
    },
    {
      "trialId": "scored:ext-008:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "ext-008",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00025760299999999997,
      "latencyMs": 8344,
      "toolCalls": 5,
      "inputTokens": 1061,
      "outputTokens": 735,
      "reasoningTokens": 0,
      "cacheReadTokens": 4608,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-008:openai/gpt-5.6-sol:2",
      "fixtureId": "ext-008",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.008624,
      "latencyMs": 9564,
      "toolCalls": 2,
      "inputTokens": 3202,
      "outputTokens": 186,
      "reasoningTokens": 36,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-008:openai/gpt-5.6-luna:2",
      "fixtureId": "ext-008",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0009710000000000001,
      "latencyMs": 9523,
      "toolCalls": 2,
      "inputTokens": 3211,
      "outputTokens": 198,
      "reasoningTokens": 76,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-008:alibaba/qwen3.8-flash:2",
      "fixtureId": "ext-008",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000466788,
      "latencyMs": 9586,
      "toolCalls": 2,
      "inputTokens": 18,
      "outputTokens": 384,
      "reasoningTokens": 146,
      "cacheReadTokens": 3013,
      "cacheWriteTokens": 833
    },
    {
      "trialId": "scored:ext-008:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "ext-008",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00518298,
      "latencyMs": 11947,
      "toolCalls": 3,
      "inputTokens": 2532,
      "outputTokens": 319,
      "reasoningTokens": 120,
      "cacheReadTokens": 2325,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-008:google/gemini-3.8-flash:2",
      "fixtureId": "ext-008",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0094905,
      "latencyMs": 14614,
      "toolCalls": 4,
      "inputTokens": 5874,
      "outputTokens": 283,
      "reasoningTokens": 1073,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-008:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "ext-008",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00747718,
      "latencyMs": 21193,
      "toolCalls": 5,
      "inputTokens": 24,
      "outputTokens": 403,
      "reasoningTokens": 244,
      "cacheReadTokens": 4704,
      "cacheWriteTokens": 1099
    },
    {
      "trialId": "scored:ext-004:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "ext-004",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000252527,
      "latencyMs": 6835,
      "toolCalls": 6,
      "inputTokens": 526,
      "outputTokens": 991,
      "reasoningTokens": 0,
      "cacheReadTokens": 4352,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:ext-008:anthropic/claude-sonnet-5:2",
      "fixtureId": "ext-008",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0112084,
      "latencyMs": 12473,
      "toolCalls": 2,
      "inputTokens": 6,
      "outputTokens": 417,
      "reasoningTokens": 94,
      "cacheReadTokens": 3532,
      "cacheWriteTokens": 2152
    },
    {
      "trialId": "scored:ext-004:openai/gpt-5.6-sol:2",
      "fixtureId": "ext-004",
      "model": "openai/gpt-5.6-sol",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0092085,
      "latencyMs": 12424,
      "toolCalls": 6,
      "inputTokens": 745,
      "outputTokens": 384,
      "reasoningTokens": 31,
      "cacheReadTokens": 2255,
      "cacheWriteTokens": 1247
    },
    {
      "trialId": "scored:ext-004:openai/gpt-5.6-luna:2",
      "fixtureId": "ext-004",
      "model": "openai/gpt-5.6-luna",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00105299,
      "latencyMs": 12456,
      "toolCalls": 6,
      "inputTokens": 745,
      "outputTokens": 384,
      "reasoningTokens": 69,
      "cacheReadTokens": 2282,
      "cacheWriteTokens": 1259
    },
    {
      "trialId": "scored:ext-004:anthropic/claude-sonnet-5:2",
      "fixtureId": "ext-004",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.017429899999999998,
      "latencyMs": 12490,
      "toolCalls": 6,
      "inputTokens": 6,
      "outputTokens": 760,
      "reasoningTokens": 150,
      "cacheReadTokens": 4377,
      "cacheWriteTokens": 2977
    },
    {
      "trialId": "scored:ext-004:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "ext-004",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.007459056,
      "latencyMs": 15637,
      "toolCalls": 7,
      "inputTokens": 3212,
      "outputTokens": 586,
      "reasoningTokens": 195,
      "cacheReadTokens": 2874,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:rec-007:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "rec-007",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01906681,
      "latencyMs": 56069,
      "toolCalls": 8,
      "inputTokens": 30,
      "outputTokens": 1114,
      "reasoningTokens": 884,
      "cacheReadTokens": 7743,
      "cacheWriteTokens": 2281
    },
    {
      "trialId": "scored:ext-004:anthropic/claude-opus-5:2",
      "fixtureId": "ext-004",
      "model": "anthropic/claude-opus-5",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.03837025,
      "latencyMs": 11861,
      "toolCalls": 6,
      "inputTokens": 6,
      "outputTokens": 832,
      "reasoningTokens": 97,
      "cacheReadTokens": 5093,
      "cacheWriteTokens": 2011
    },
    {
      "trialId": "scored:ext-004:alibaba/qwen3.8-flash:2",
      "fixtureId": "ext-004",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.000929724,
      "latencyMs": 14037,
      "toolCalls": 6,
      "inputTokens": 18,
      "outputTokens": 781,
      "reasoningTokens": 437,
      "cacheReadTokens": 3424,
      "cacheWriteTokens": 1498
    },
    {
      "trialId": "scored:ext-004:google/gemini-3.8-flash:2",
      "fixtureId": "ext-004",
      "model": "google/gemini-3.8-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.013610249999999999,
      "latencyMs": 17153,
      "toolCalls": 7,
      "inputTokens": 7287,
      "outputTokens": 359,
      "reasoningTokens": 1813,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-007:deepseek/deepseek-v4-pro-0813:2",
      "fixtureId": "sup-007",
      "model": "deepseek/deepseek-v4-pro-0813",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.007976716,
      "latencyMs": 18932,
      "toolCalls": 7,
      "inputTokens": 3429,
      "outputTokens": 650,
      "reasoningTokens": 186,
      "cacheReadTokens": 3179,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-007:deepseek/deepseek-v4-flash-0731:2",
      "fixtureId": "sup-007",
      "model": "deepseek/deepseek-v4-flash-0731",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00029281500000000003,
      "latencyMs": 7079,
      "toolCalls": 7,
      "inputTokens": 986,
      "outputTokens": 815,
      "reasoningTokens": 0,
      "cacheReadTokens": 6656,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-007:openai/gpt-5.6-sol:2",
      "fixtureId": "sup-007",
      "model": "openai/gpt-5.6-sol",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.009146300000000001,
      "latencyMs": 13146,
      "toolCalls": 6,
      "inputTokens": 795,
      "outputTokens": 303,
      "reasoningTokens": 44,
      "cacheReadTokens": 2669,
      "cacheWriteTokens": 1421
    },
    {
      "trialId": "scored:ext-004:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "ext-004",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.011744499999999998,
      "latencyMs": 29008,
      "toolCalls": 7,
      "inputTokens": 18,
      "outputTokens": 933,
      "reasoningTokens": 223,
      "cacheReadTokens": 3500,
      "cacheWriteTokens": 1671
    },
    {
      "trialId": "scored:sup-007:google/gemini-3.8-flash:2",
      "fixtureId": "sup-007",
      "model": "google/gemini-3.8-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0134055,
      "latencyMs": 17524,
      "toolCalls": 8,
      "inputTokens": 10134,
      "outputTokens": 385,
      "reasoningTokens": 1163,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "scored:sup-007:alibaba/qwen3.8-flash:2",
      "fixtureId": "sup-007",
      "model": "alibaba/qwen3.8-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.001169366,
      "latencyMs": 18754,
      "toolCalls": 10,
      "inputTokens": 36,
      "outputTokens": 819,
      "reasoningTokens": 450,
      "cacheReadTokens": 10036,
      "cacheWriteTokens": 2033
    },
    {
      "trialId": "scored:sup-007:openai/gpt-5.6-luna:2",
      "fixtureId": "sup-007",
      "model": "openai/gpt-5.6-luna",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00118516,
      "latencyMs": 9667,
      "toolCalls": 6,
      "inputTokens": 795,
      "outputTokens": 288,
      "reasoningTokens": 197,
      "cacheReadTokens": 2858,
      "cacheWriteTokens": 1548
    },
    {
      "trialId": "scored:sup-007:alibaba/qwen3.8-max-0902:2",
      "fixtureId": "sup-007",
      "model": "alibaba/qwen3.8-max-0902",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.01062134,
      "latencyMs": 27961,
      "toolCalls": 8,
      "inputTokens": 24,
      "outputTokens": 680,
      "reasoningTokens": 231,
      "cacheReadTokens": 5852,
      "cacheWriteTokens": 1645
    },
    {
      "trialId": "scored:sup-007:anthropic/claude-opus-5:2",
      "fixtureId": "sup-007",
      "model": "anthropic/claude-opus-5",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.043266,
      "latencyMs": 14619,
      "toolCalls": 8,
      "inputTokens": 8,
      "outputTokens": 815,
      "reasoningTokens": 127,
      "cacheReadTokens": 8552,
      "cacheWriteTokens": 2464
    },
    {
      "trialId": "scored:sup-007:anthropic/claude-sonnet-5:2",
      "fixtureId": "sup-007",
      "model": "anthropic/claude-sonnet-5",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0195198,
      "latencyMs": 15009,
      "toolCalls": 6,
      "inputTokens": 8,
      "outputTokens": 802,
      "reasoningTokens": 188,
      "cacheReadTokens": 7644,
      "cacheWriteTokens": 3230
    },
    {
      "trialId": "extension-v1:scored:ext-003:zai/glm-5.3-flash:0",
      "fixtureId": "ext-003",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0007393,
      "latencyMs": 4750,
      "toolCalls": 2,
      "inputTokens": 4012,
      "outputTokens": 244,
      "reasoningTokens": 31,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-003:zai/glm-5.3-flash:0",
      "fixtureId": "rec-003",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00082017,
      "latencyMs": 5904,
      "toolCalls": 7,
      "inputTokens": 3973,
      "outputTokens": 387,
      "reasoningTokens": 0,
      "cacheReadTokens": 1024,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-004:zai/glm-5.3-flash:0",
      "fixtureId": "sup-004",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0011346,
      "latencyMs": 8684,
      "toolCalls": 6,
      "inputTokens": 5814,
      "outputTokens": 469,
      "reasoningTokens": 56,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-001:zai/glm-5.3-flash:0",
      "fixtureId": "rec-001",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00071435,
      "latencyMs": 11616,
      "toolCalls": 4,
      "inputTokens": 4049,
      "outputTokens": 214,
      "reasoningTokens": 0,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-003:openai/gpt-6-astra:0",
      "fixtureId": "ext-003",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.038363999999999995,
      "latencyMs": 11797,
      "toolCalls": 2,
      "inputTokens": 951,
      "outputTokens": 230,
      "reasoningTokens": 0,
      "cacheReadTokens": 2254,
      "cacheWriteTokens": 1208
    },
    {
      "trialId": "extension-v1:scored:rec-001:openai/gpt-6-astra:0",
      "fixtureId": "rec-001",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.042910000000000004,
      "latencyMs": 12323,
      "toolCalls": 3,
      "inputTokens": 3196,
      "outputTokens": 219,
      "reasoningTokens": 0,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-003:openai/gpt-6-astra:0",
      "fixtureId": "rec-003",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0424635,
      "latencyMs": 14217,
      "toolCalls": 5,
      "inputTokens": 753,
      "outputTokens": 348,
      "reasoningTokens": 0,
      "cacheReadTokens": 2246,
      "cacheWriteTokens": 1223
    },
    {
      "trialId": "extension-v1:scored:sup-004:openai/gpt-6-astra:0",
      "fixtureId": "sup-004",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.046371,
      "latencyMs": 15877,
      "toolCalls": 5,
      "inputTokens": 844,
      "outputTokens": 350,
      "reasoningTokens": 0,
      "cacheReadTokens": 2631,
      "cacheWriteTokens": 1424
    },
    {
      "trialId": "extension-v1:scored:ext-007:zai/glm-5.3-flash:0",
      "fixtureId": "ext-007",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00073695,
      "latencyMs": 8863,
      "toolCalls": 4,
      "inputTokens": 3723,
      "outputTokens": 321,
      "reasoningTokens": 36,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-007:openai/gpt-6-astra:0",
      "fixtureId": "ext-007",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.03991,
      "latencyMs": 16978,
      "toolCalls": 4,
      "inputTokens": 837,
      "outputTokens": 292,
      "reasoningTokens": 0,
      "cacheReadTokens": 2190,
      "cacheWriteTokens": 1180
    },
    {
      "trialId": "extension-v1:scored:rec-002:zai/glm-5.3-flash:0",
      "fixtureId": "rec-002",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00085425,
      "latencyMs": 7420,
      "toolCalls": 5,
      "inputTokens": 4545,
      "outputTokens": 322,
      "reasoningTokens": 23,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-002:zai/glm-5.3-flash:0",
      "fixtureId": "sup-002",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0007642,
      "latencyMs": 7459,
      "toolCalls": 4,
      "inputTokens": 3868,
      "outputTokens": 328,
      "reasoningTokens": 40,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-002:openai/gpt-6-astra:0",
      "fixtureId": "sup-002",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.041272500000000004,
      "latencyMs": 17703,
      "toolCalls": 4,
      "inputTokens": 803,
      "outputTokens": 311,
      "reasoningTokens": 0,
      "cacheReadTokens": 2280,
      "cacheWriteTokens": 1233
    },
    {
      "trialId": "extension-v1:scored:ext-010:zai/glm-5.3-flash:0",
      "fixtureId": "ext-010",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0010002499999999998,
      "latencyMs": 11547,
      "toolCalls": 4,
      "inputTokens": 4935,
      "outputTokens": 382,
      "reasoningTokens": 138,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-002:openai/gpt-6-astra:0",
      "fixtureId": "rec-002",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.051745,
      "latencyMs": 17869,
      "toolCalls": 4,
      "inputTokens": 2607,
      "outputTokens": 255,
      "reasoningTokens": 0,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 1034
    },
    {
      "trialId": "extension-v1:scored:ext-010:openai/gpt-6-astra:0",
      "fixtureId": "ext-010",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.053556000000000006,
      "latencyMs": 17886,
      "toolCalls": 4,
      "inputTokens": 2461,
      "outputTokens": 292,
      "reasoningTokens": 0,
      "cacheReadTokens": 1046,
      "cacheWriteTokens": 1064
    },
    {
      "trialId": "extension-v1:scored:ext-001:openai/gpt-6-astra:0",
      "fixtureId": "ext-001",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.043809,
      "latencyMs": 11583,
      "toolCalls": 2,
      "inputTokens": 883,
      "outputTokens": 329,
      "reasoningTokens": 0,
      "cacheReadTokens": 2379,
      "cacheWriteTokens": 1292
    },
    {
      "trialId": "extension-v1:scored:rec-010:zai/glm-5.3-flash:0",
      "fixtureId": "rec-010",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0006758000000000001,
      "latencyMs": 6603,
      "toolCalls": 6,
      "inputTokens": 3422,
      "outputTokens": 278,
      "reasoningTokens": 47,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-001:zai/glm-5.3-flash:0",
      "fixtureId": "ext-001",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0008053000000000001,
      "latencyMs": 7051,
      "toolCalls": 2,
      "inputTokens": 4112,
      "outputTokens": 307,
      "reasoningTokens": 70,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-010:openai/gpt-6-astra:0",
      "fixtureId": "rec-010",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.045233499999999996,
      "latencyMs": 12374,
      "toolCalls": 6,
      "inputTokens": 1648,
      "outputTokens": 276,
      "reasoningTokens": 0,
      "cacheReadTokens": 1091,
      "cacheWriteTokens": 1109
    },
    {
      "trialId": "extension-v1:scored:ext-006:zai/glm-5.3-flash:0",
      "fixtureId": "ext-006",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00079855,
      "latencyMs": 3850,
      "toolCalls": 5,
      "inputTokens": 3771,
      "outputTokens": 348,
      "reasoningTokens": 41,
      "cacheReadTokens": 1280,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-005:zai/glm-5.3-flash:0",
      "fixtureId": "rec-005",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0005217,
      "latencyMs": 10920,
      "toolCalls": 3,
      "inputTokens": 2858,
      "outputTokens": 168,
      "reasoningTokens": 18,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-005:openai/gpt-6-astra:0",
      "fixtureId": "rec-005",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.03905,
      "latencyMs": 11262,
      "toolCalls": 3,
      "inputTokens": 2975,
      "outputTokens": 186,
      "reasoningTokens": 0,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-001:openai/gpt-6-astra:0",
      "fixtureId": "sup-001",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.042431,
      "latencyMs": 12343,
      "toolCalls": 5,
      "inputTokens": 750,
      "outputTokens": 330,
      "reasoningTokens": 0,
      "cacheReadTokens": 2381,
      "cacheWriteTokens": 1284
    },
    {
      "trialId": "extension-v1:scored:sup-010:openai/gpt-6-astra:0",
      "fixtureId": "sup-010",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0459175,
      "latencyMs": 12961,
      "toolCalls": 4,
      "inputTokens": 904,
      "outputTokens": 328,
      "reasoningTokens": 0,
      "cacheReadTokens": 2640,
      "cacheWriteTokens": 1427
    },
    {
      "trialId": "extension-v1:scored:sup-010:zai/glm-5.3-flash:0",
      "fixtureId": "sup-010",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0009245,
      "latencyMs": 13558,
      "toolCalls": 4,
      "inputTokens": 4390,
      "outputTokens": 484,
      "reasoningTokens": 48,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-009:zai/glm-5.3-flash:0",
      "fixtureId": "sup-009",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0010010800000000001,
      "latencyMs": 5782,
      "toolCalls": 6,
      "inputTokens": 4586,
      "outputTokens": 541,
      "reasoningTokens": 70,
      "cacheReadTokens": 256,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-006:openai/gpt-6-astra:0",
      "fixtureId": "ext-006",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.043828500000000006,
      "latencyMs": 18220,
      "toolCalls": 5,
      "inputTokens": 754,
      "outputTokens": 348,
      "reasoningTokens": 0,
      "cacheReadTokens": 3351,
      "cacheWriteTokens": 1243
    },
    {
      "trialId": "extension-v1:scored:ext-004:zai/glm-5.3-flash:0",
      "fixtureId": "ext-004",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00084775,
      "latencyMs": 10700,
      "toolCalls": 6,
      "inputTokens": 3865,
      "outputTokens": 459,
      "reasoningTokens": 77,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-003:openai/gpt-6-astra:0",
      "fixtureId": "sup-003",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.041528,
      "latencyMs": 13103,
      "toolCalls": 4,
      "inputTokens": 846,
      "outputTokens": 293,
      "reasoningTokens": 0,
      "cacheReadTokens": 2393,
      "cacheWriteTokens": 1282
    },
    {
      "trialId": "extension-v1:scored:ext-004:openai/gpt-6-astra:0",
      "fixtureId": "ext-004",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.057701,
      "latencyMs": 14710,
      "toolCalls": 6,
      "inputTokens": 745,
      "outputTokens": 382,
      "reasoningTokens": 0,
      "cacheReadTokens": 1026,
      "cacheWriteTokens": 2410
    },
    {
      "trialId": "extension-v1:scored:sup-009:openai/gpt-6-astra:0",
      "fixtureId": "sup-009",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0550505,
      "latencyMs": 15673,
      "toolCalls": 6,
      "inputTokens": 892,
      "outputTokens": 456,
      "reasoningTokens": 0,
      "cacheReadTokens": 2993,
      "cacheWriteTokens": 1627
    },
    {
      "trialId": "extension-v1:scored:sup-001:zai/glm-5.3-flash:0",
      "fixtureId": "sup-001",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00091805,
      "latencyMs": 32682,
      "toolCalls": 5,
      "inputTokens": 4057,
      "outputTokens": 544,
      "reasoningTokens": 75,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-007:zai/glm-5.3-flash:0",
      "fixtureId": "rec-007",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0008851,
      "latencyMs": 5910,
      "toolCalls": 6,
      "inputTokens": 4814,
      "outputTokens": 326,
      "reasoningTokens": 0,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-008:zai/glm-5.3-flash:0",
      "fixtureId": "rec-008",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00073343,
      "latencyMs": 6628,
      "toolCalls": 4,
      "inputTokens": 3355,
      "outputTokens": 390,
      "reasoningTokens": 55,
      "cacheReadTokens": 256,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-003:zai/glm-5.3-flash:0",
      "fixtureId": "sup-003",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0009219,
      "latencyMs": 24081,
      "toolCalls": 4,
      "inputTokens": 4256,
      "outputTokens": 418,
      "reasoningTokens": 149,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-008:openai/gpt-6-astra:0",
      "fixtureId": "rec-008",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0462955,
      "latencyMs": 13133,
      "toolCalls": 4,
      "inputTokens": 1652,
      "outputTokens": 304,
      "reasoningTokens": 0,
      "cacheReadTokens": 1063,
      "cacheWriteTokens": 1081
    },
    {
      "trialId": "extension-v1:scored:rec-009:zai/glm-5.3-flash:0",
      "fixtureId": "rec-009",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00088513,
      "latencyMs": 14543,
      "toolCalls": 4,
      "inputTokens": 3823,
      "outputTokens": 403,
      "reasoningTokens": 205,
      "cacheReadTokens": 256,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-006:zai/glm-5.3-flash:0",
      "fixtureId": "rec-006",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0006628000000000001,
      "latencyMs": 9667,
      "toolCalls": 4,
      "inputTokens": 3262,
      "outputTokens": 291,
      "reasoningTokens": 56,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-006:openai/gpt-6-astra:0",
      "fixtureId": "rec-006",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.04853,
      "latencyMs": 12222,
      "toolCalls": 4,
      "inputTokens": 3523,
      "outputTokens": 266,
      "reasoningTokens": 0,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-009:openai/gpt-6-astra:0",
      "fixtureId": "rec-009",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0436125,
      "latencyMs": 12852,
      "toolCalls": 4,
      "inputTokens": 885,
      "outputTokens": 340,
      "reasoningTokens": 0,
      "cacheReadTokens": 2300,
      "cacheWriteTokens": 1237
    },
    {
      "trialId": "extension-v1:scored:rec-004:zai/glm-5.3-flash:0",
      "fixtureId": "rec-004",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00069935,
      "latencyMs": 8475,
      "toolCalls": 4,
      "inputTokens": 3679,
      "outputTokens": 283,
      "reasoningTokens": 12,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-007:openai/gpt-6-astra:0",
      "fixtureId": "rec-007",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.06660600000000001,
      "latencyMs": 19867,
      "toolCalls": 6,
      "inputTokens": 1737,
      "outputTokens": 366,
      "reasoningTokens": 0,
      "cacheReadTokens": 1036,
      "cacheWriteTokens": 2392
    },
    {
      "trialId": "extension-v1:scored:ext-002:zai/glm-5.3-flash:0",
      "fixtureId": "ext-002",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0006954800000000001,
      "latencyMs": 5256,
      "toolCalls": 3,
      "inputTokens": 3492,
      "outputTokens": 311,
      "reasoningTokens": 17,
      "cacheReadTokens": 256,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-004:openai/gpt-6-astra:0",
      "fixtureId": "rec-004",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.041499,
      "latencyMs": 15519,
      "toolCalls": 4,
      "inputTokens": 782,
      "outputTokens": 331,
      "reasoningTokens": 0,
      "cacheReadTokens": 2204,
      "cacheWriteTokens": 1194
    },
    {
      "trialId": "extension-v1:scored:ext-009:zai/glm-5.3-flash:0",
      "fixtureId": "ext-009",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0005235600000000001,
      "latencyMs": 8470,
      "toolCalls": 2,
      "inputTokens": 2618,
      "outputTokens": 231,
      "reasoningTokens": 0,
      "cacheReadTokens": 512,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-009:openai/gpt-6-astra:0",
      "fixtureId": "ext-009",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0431,
      "latencyMs": 11497,
      "toolCalls": 2,
      "inputTokens": 3295,
      "outputTokens": 203,
      "reasoningTokens": 0,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-005:zai/glm-5.3-flash:0",
      "fixtureId": "sup-005",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00104468,
      "latencyMs": 12601,
      "toolCalls": 5,
      "inputTokens": 4600,
      "outputTokens": 453,
      "reasoningTokens": 241,
      "cacheReadTokens": 256,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-002:openai/gpt-6-astra:0",
      "fixtureId": "ext-002",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.042773,
      "latencyMs": 13037,
      "toolCalls": 3,
      "inputTokens": 785,
      "outputTokens": 348,
      "reasoningTokens": 0,
      "cacheReadTokens": 2248,
      "cacheWriteTokens": 1222
    },
    {
      "trialId": "extension-v1:scored:sup-007:openai/gpt-6-astra:0",
      "fixtureId": "sup-007",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.062858,
      "latencyMs": 17646,
      "toolCalls": 7,
      "inputTokens": 798,
      "outputTokens": 341,
      "reasoningTokens": 0,
      "cacheReadTokens": 2553,
      "cacheWriteTokens": 2822
    },
    {
      "trialId": "extension-v1:scored:ext-005:zai/glm-5.3-flash:0",
      "fixtureId": "ext-005",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00068895,
      "latencyMs": 6324,
      "toolCalls": 4,
      "inputTokens": 3333,
      "outputTokens": 238,
      "reasoningTokens": 140,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-008:zai/glm-5.3-flash:0",
      "fixtureId": "sup-008",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0011596100000000002,
      "latencyMs": 8869,
      "toolCalls": 7,
      "inputTokens": 5045,
      "outputTokens": 668,
      "reasoningTokens": 107,
      "cacheReadTokens": 512,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-007:zai/glm-5.3-flash:0",
      "fixtureId": "sup-007",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0006547,
      "latencyMs": 9784,
      "toolCalls": 6,
      "inputTokens": 2832,
      "outputTokens": 327,
      "reasoningTokens": 56,
      "cacheReadTokens": 1280,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-005:openai/gpt-6-astra:0",
      "fixtureId": "sup-005",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0519245,
      "latencyMs": 22238,
      "toolCalls": 6,
      "inputTokens": 962,
      "outputTokens": 375,
      "reasoningTokens": 0,
      "cacheReadTokens": 4267,
      "cacheWriteTokens": 1543
    },
    {
      "trialId": "extension-v1:scored:sup-006:openai/gpt-6-astra:0",
      "fixtureId": "sup-006",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.03969899999999999,
      "latencyMs": 14299,
      "toolCalls": 5,
      "inputTokens": 836,
      "outputTokens": 248,
      "reasoningTokens": 0,
      "cacheReadTokens": 2489,
      "cacheWriteTokens": 1316
    },
    {
      "trialId": "extension-v1:scored:sup-008:openai/gpt-6-astra:0",
      "fixtureId": "sup-008",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0592635,
      "latencyMs": 21216,
      "toolCalls": 7,
      "inputTokens": 12,
      "outputTokens": 600,
      "reasoningTokens": 0,
      "cacheReadTokens": 4606,
      "cacheWriteTokens": 1963
    },
    {
      "trialId": "extension-v1:scored:ext-008:zai/glm-5.3-flash:1",
      "fixtureId": "ext-008",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00062688,
      "latencyMs": 7255,
      "toolCalls": 2,
      "inputTokens": 2938,
      "outputTokens": 293,
      "reasoningTokens": 64,
      "cacheReadTokens": 256,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-006:zai/glm-5.3-flash:0",
      "fixtureId": "sup-006",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00084763,
      "latencyMs": 14810,
      "toolCalls": 5,
      "inputTokens": 3953,
      "outputTokens": 438,
      "reasoningTokens": 56,
      "cacheReadTokens": 256,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-005:openai/gpt-6-astra:0",
      "fixtureId": "ext-005",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.04731,
      "latencyMs": 14819,
      "toolCalls": 4,
      "inputTokens": 3516,
      "outputTokens": 243,
      "reasoningTokens": 0,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-008:zai/glm-5.3-flash:0",
      "fixtureId": "ext-008",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00056534,
      "latencyMs": 7518,
      "toolCalls": 2,
      "inputTokens": 2342,
      "outputTokens": 340,
      "reasoningTokens": 42,
      "cacheReadTokens": 768,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-008:openai/gpt-6-astra:0",
      "fixtureId": "ext-008",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 0,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.06035,
      "latencyMs": 18198,
      "toolCalls": 4,
      "inputTokens": 4835,
      "outputTokens": 240,
      "reasoningTokens": 0,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-008:openai/gpt-6-astra:1",
      "fixtureId": "ext-008",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.05114,
      "latencyMs": 14064,
      "toolCalls": 3,
      "inputTokens": 4004,
      "outputTokens": 222,
      "reasoningTokens": 0,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-002:zai/glm-5.3-flash:1",
      "fixtureId": "rec-002",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0007096700000000001,
      "latencyMs": 9801,
      "toolCalls": 5,
      "inputTokens": 3513,
      "outputTokens": 285,
      "reasoningTokens": 19,
      "cacheReadTokens": 1024,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-010:zai/glm-5.3-flash:1",
      "fixtureId": "sup-010",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0008607300000000001,
      "latencyMs": 11657,
      "toolCalls": 4,
      "inputTokens": 4077,
      "outputTokens": 440,
      "reasoningTokens": 43,
      "cacheReadTokens": 256,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-002:openai/gpt-6-astra:1",
      "fixtureId": "rec-002",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.039903999999999995,
      "latencyMs": 12319,
      "toolCalls": 4,
      "inputTokens": 2607,
      "outputTokens": 256,
      "reasoningTokens": 0,
      "cacheReadTokens": 1034,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-004:openai/gpt-6-astra:1",
      "fixtureId": "sup-004",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.03278349999999999,
      "latencyMs": 12903,
      "toolCalls": 5,
      "inputTokens": 844,
      "outputTokens": 360,
      "reasoningTokens": 0,
      "cacheReadTokens": 3856,
      "cacheWriteTokens": 199
    },
    {
      "trialId": "extension-v1:scored:sup-010:openai/gpt-6-astra:1",
      "fixtureId": "sup-010",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0328785,
      "latencyMs": 13112,
      "toolCalls": 4,
      "inputTokens": 904,
      "outputTokens": 349,
      "reasoningTokens": 0,
      "cacheReadTokens": 3876,
      "cacheWriteTokens": 201
    },
    {
      "trialId": "extension-v1:scored:sup-007:zai/glm-5.3-flash:1",
      "fixtureId": "sup-007",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": false,
      "reasons": [
        "submission.params: unexpected property customer"
      ],
      "cost": 0.0008389599999999999,
      "latencyMs": 6875,
      "toolCalls": 6,
      "inputTokens": 3744,
      "outputTokens": 454,
      "reasoningTokens": 70,
      "cacheReadTokens": 512,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-005:zai/glm-5.3-flash:1",
      "fixtureId": "rec-005",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00060321,
      "latencyMs": 7492,
      "toolCalls": 4,
      "inputTokens": 3329,
      "outputTokens": 177,
      "reasoningTokens": 0,
      "cacheReadTokens": 512,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-007:zai/glm-5.3-flash:1",
      "fixtureId": "rec-007",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00087218,
      "latencyMs": 14131,
      "toolCalls": 6,
      "inputTokens": 4610,
      "outputTokens": 322,
      "reasoningTokens": 24,
      "cacheReadTokens": 256,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-005:openai/gpt-6-astra:1",
      "fixtureId": "sup-005",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.03393,
      "latencyMs": 14906,
      "toolCalls": 6,
      "inputTokens": 962,
      "outputTokens": 370,
      "reasoningTokens": 0,
      "cacheReadTokens": 5810,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-004:zai/glm-5.3-flash:1",
      "fixtureId": "sup-004",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00078979,
      "latencyMs": 29654,
      "toolCalls": 5,
      "inputTokens": 3545,
      "outputTokens": 419,
      "reasoningTokens": 51,
      "cacheReadTokens": 768,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-007:openai/gpt-6-astra:1",
      "fixtureId": "rec-007",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0521345,
      "latencyMs": 16181,
      "toolCalls": 6,
      "inputTokens": 1737,
      "outputTokens": 349,
      "reasoningTokens": 0,
      "cacheReadTokens": 2227,
      "cacheWriteTokens": 1207
    },
    {
      "trialId": "extension-v1:scored:sup-007:openai/gpt-6-astra:1",
      "fixtureId": "sup-007",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.031469000000000004,
      "latencyMs": 16409,
      "toolCalls": 7,
      "inputTokens": 798,
      "outputTokens": 323,
      "reasoningTokens": 0,
      "cacheReadTokens": 5214,
      "cacheWriteTokens": 170
    },
    {
      "trialId": "extension-v1:scored:ext-009:zai/glm-5.3-flash:1",
      "fixtureId": "ext-009",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00053756,
      "latencyMs": 6890,
      "toolCalls": 2,
      "inputTokens": 2618,
      "outputTokens": 259,
      "reasoningTokens": 0,
      "cacheReadTokens": 512,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-005:zai/glm-5.3-flash:1",
      "fixtureId": "sup-005",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00144569,
      "latencyMs": 40094,
      "toolCalls": 6,
      "inputTokens": 6311,
      "outputTokens": 546,
      "reasoningTokens": 406,
      "cacheReadTokens": 768,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-009:openai/gpt-6-astra:1",
      "fixtureId": "ext-009",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0433,
      "latencyMs": 13947,
      "toolCalls": 2,
      "inputTokens": 3295,
      "outputTokens": 207,
      "reasoningTokens": 0,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-010:zai/glm-5.3-flash:1",
      "fixtureId": "ext-010",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00073778,
      "latencyMs": 6760,
      "toolCalls": 3,
      "inputTokens": 3264,
      "outputTokens": 351,
      "reasoningTokens": 130,
      "cacheReadTokens": 256,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-005:openai/gpt-6-astra:1",
      "fixtureId": "rec-005",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.03905,
      "latencyMs": 14102,
      "toolCalls": 3,
      "inputTokens": 2975,
      "outputTokens": 186,
      "reasoningTokens": 0,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-008:openai/gpt-6-astra:1",
      "fixtureId": "rec-008",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.033564,
      "latencyMs": 17087,
      "toolCalls": 4,
      "inputTokens": 1652,
      "outputTokens": 298,
      "reasoningTokens": 0,
      "cacheReadTokens": 2144,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-008:openai/gpt-6-astra:1",
      "fixtureId": "sup-008",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0282295,
      "latencyMs": 13422,
      "toolCalls": 7,
      "inputTokens": 9,
      "outputTokens": 406,
      "reasoningTokens": 0,
      "cacheReadTokens": 4277,
      "cacheWriteTokens": 285
    },
    {
      "trialId": "extension-v1:scored:rec-008:zai/glm-5.3-flash:1",
      "fixtureId": "rec-008",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0006923000000000001,
      "latencyMs": 14332,
      "toolCalls": 4,
      "inputTokens": 3442,
      "outputTokens": 315,
      "reasoningTokens": 37,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-010:openai/gpt-6-astra:1",
      "fixtureId": "ext-010",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.04142,
      "latencyMs": 17605,
      "toolCalls": 4,
      "inputTokens": 2461,
      "outputTokens": 294,
      "reasoningTokens": 0,
      "cacheReadTokens": 2110,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-010:zai/glm-5.3-flash:1",
      "fixtureId": "rec-010",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00062069,
      "latencyMs": 6760,
      "toolCalls": 6,
      "inputTokens": 2761,
      "outputTokens": 320,
      "reasoningTokens": 47,
      "cacheReadTokens": 768,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-009:zai/glm-5.3-flash:1",
      "fixtureId": "sup-009",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00102158,
      "latencyMs": 8704,
      "toolCalls": 6,
      "inputTokens": 4616,
      "outputTokens": 567,
      "reasoningTokens": 76,
      "cacheReadTokens": 256,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-010:openai/gpt-6-astra:1",
      "fixtureId": "rec-010",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0465835,
      "latencyMs": 11177,
      "toolCalls": 6,
      "inputTokens": 1648,
      "outputTokens": 303,
      "reasoningTokens": 0,
      "cacheReadTokens": 1091,
      "cacheWriteTokens": 1109
    },
    {
      "trialId": "extension-v1:scored:sup-008:zai/glm-5.3-flash:1",
      "fixtureId": "sup-008",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00121973,
      "latencyMs": 15722,
      "toolCalls": 7,
      "inputTokens": 5347,
      "outputTokens": 751,
      "reasoningTokens": 69,
      "cacheReadTokens": 256,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-007:zai/glm-5.3-flash:1",
      "fixtureId": "ext-007",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00064261,
      "latencyMs": 6637,
      "toolCalls": 4,
      "inputTokens": 3175,
      "outputTokens": 302,
      "reasoningTokens": 0,
      "cacheReadTokens": 512,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-009:openai/gpt-6-astra:1",
      "fixtureId": "sup-009",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.037763,
      "latencyMs": 19748,
      "toolCalls": 6,
      "inputTokens": 892,
      "outputTokens": 431,
      "reasoningTokens": 0,
      "cacheReadTokens": 4368,
      "cacheWriteTokens": 234
    },
    {
      "trialId": "extension-v1:scored:ext-001:zai/glm-5.3-flash:1",
      "fixtureId": "ext-001",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00077406,
      "latencyMs": 10211,
      "toolCalls": 2,
      "inputTokens": 3578,
      "outputTokens": 396,
      "reasoningTokens": 48,
      "cacheReadTokens": 512,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-005:openai/gpt-6-astra:1",
      "fixtureId": "ext-005",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.04731,
      "latencyMs": 10477,
      "toolCalls": 4,
      "inputTokens": 3516,
      "outputTokens": 243,
      "reasoningTokens": 0,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-005:zai/glm-5.3-flash:1",
      "fixtureId": "ext-005",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00061163,
      "latencyMs": 14042,
      "toolCalls": 4,
      "inputTokens": 2983,
      "outputTokens": 269,
      "reasoningTokens": 44,
      "cacheReadTokens": 256,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-007:openai/gpt-6-astra:1",
      "fixtureId": "ext-007",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.026090000000000002,
      "latencyMs": 14083,
      "toolCalls": 4,
      "inputTokens": 837,
      "outputTokens": 287,
      "reasoningTokens": 0,
      "cacheReadTokens": 3370,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-004:zai/glm-5.3-flash:1",
      "fixtureId": "ext-004",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0008315099999999999,
      "latencyMs": 9226,
      "toolCalls": 6,
      "inputTokens": 3421,
      "outputTokens": 512,
      "reasoningTokens": 94,
      "cacheReadTokens": 512,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-003:zai/glm-5.3-flash:1",
      "fixtureId": "ext-003",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00068714,
      "latencyMs": 4512,
      "toolCalls": 2,
      "inputTokens": 3284,
      "outputTokens": 329,
      "reasoningTokens": 14,
      "cacheReadTokens": 768,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-001:openai/gpt-6-astra:1",
      "fixtureId": "ext-001",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.028200999999999997,
      "latencyMs": 13121,
      "toolCalls": 2,
      "inputTokens": 883,
      "outputTokens": 314,
      "reasoningTokens": 0,
      "cacheReadTokens": 3671,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-004:openai/gpt-6-astra:1",
      "fixtureId": "ext-004",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.029886,
      "latencyMs": 16888,
      "toolCalls": 6,
      "inputTokens": 745,
      "outputTokens": 380,
      "reasoningTokens": 0,
      "cacheReadTokens": 3436,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-004:openai/gpt-6-astra:1",
      "fixtureId": "rec-004",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.027768,
      "latencyMs": 12763,
      "toolCalls": 4,
      "inputTokens": 782,
      "outputTokens": 331,
      "reasoningTokens": 0,
      "cacheReadTokens": 3398,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-001:openai/gpt-6-astra:1",
      "fixtureId": "rec-001",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.042710000000000005,
      "latencyMs": 13233,
      "toolCalls": 3,
      "inputTokens": 3196,
      "outputTokens": 215,
      "reasoningTokens": 0,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-004:zai/glm-5.3-flash:1",
      "fixtureId": "rec-004",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00063061,
      "latencyMs": 5575,
      "toolCalls": 4,
      "inputTokens": 3165,
      "outputTokens": 276,
      "reasoningTokens": 5,
      "cacheReadTokens": 512,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-003:openai/gpt-6-astra:1",
      "fixtureId": "ext-003",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.024722,
      "latencyMs": 14690,
      "toolCalls": 2,
      "inputTokens": 951,
      "outputTokens": 235,
      "reasoningTokens": 0,
      "cacheReadTokens": 3462,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-001:zai/glm-5.3-flash:1",
      "fixtureId": "rec-001",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0005593099999999999,
      "latencyMs": 19891,
      "toolCalls": 3,
      "inputTokens": 2573,
      "outputTokens": 271,
      "reasoningTokens": 45,
      "cacheReadTokens": 512,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-002:zai/glm-5.3-flash:1",
      "fixtureId": "ext-002",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00068956,
      "latencyMs": 7604,
      "toolCalls": 3,
      "inputTokens": 3238,
      "outputTokens": 358,
      "reasoningTokens": 19,
      "cacheReadTokens": 512,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-002:openai/gpt-6-astra:1",
      "fixtureId": "ext-002",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.020368,
      "latencyMs": 9122,
      "toolCalls": 3,
      "inputTokens": 782,
      "outputTokens": 206,
      "reasoningTokens": 0,
      "cacheReadTokens": 2248,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-003:openai/gpt-6-astra:1",
      "fixtureId": "sup-003",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.028408,
      "latencyMs": 17243,
      "toolCalls": 4,
      "inputTokens": 846,
      "outputTokens": 290,
      "reasoningTokens": 0,
      "cacheReadTokens": 3523,
      "cacheWriteTokens": 154
    },
    {
      "trialId": "extension-v1:scored:rec-003:openai/gpt-6-astra:1",
      "fixtureId": "rec-003",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.027849,
      "latencyMs": 11633,
      "toolCalls": 5,
      "inputTokens": 753,
      "outputTokens": 337,
      "reasoningTokens": 0,
      "cacheReadTokens": 3469,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-003:zai/glm-5.3-flash:1",
      "fixtureId": "rec-003",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00098802,
      "latencyMs": 13936,
      "toolCalls": 7,
      "inputTokens": 5222,
      "outputTokens": 348,
      "reasoningTokens": 0,
      "cacheReadTokens": 1024,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-003:zai/glm-5.3-flash:1",
      "fixtureId": "sup-003",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00081058,
      "latencyMs": 21625,
      "toolCalls": 4,
      "inputTokens": 3816,
      "outputTokens": 416,
      "reasoningTokens": 45,
      "cacheReadTokens": 256,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-006:zai/glm-5.3-flash:1",
      "fixtureId": "sup-006",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0008429799999999999,
      "latencyMs": 16716,
      "toolCalls": 5,
      "inputTokens": 4002,
      "outputTokens": 401,
      "reasoningTokens": 69,
      "cacheReadTokens": 256,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-006:zai/glm-5.3-flash:1",
      "fixtureId": "ext-006",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0006293900000000001,
      "latencyMs": 12543,
      "toolCalls": 4,
      "inputTokens": 2929,
      "outputTokens": 292,
      "reasoningTokens": 42,
      "cacheReadTokens": 768,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-006:openai/gpt-6-astra:1",
      "fixtureId": "sup-006",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.040092,
      "latencyMs": 12552,
      "toolCalls": 5,
      "inputTokens": 836,
      "outputTokens": 260,
      "reasoningTokens": 0,
      "cacheReadTokens": 2507,
      "cacheWriteTokens": 1298
    },
    {
      "trialId": "extension-v1:scored:rec-006:zai/glm-5.3-flash:1",
      "fixtureId": "rec-006",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00065608,
      "latencyMs": 13072,
      "toolCalls": 4,
      "inputTokens": 3126,
      "outputTokens": 316,
      "reasoningTokens": 43,
      "cacheReadTokens": 256,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-006:openai/gpt-6-astra:1",
      "fixtureId": "ext-006",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.028631499999999997,
      "latencyMs": 13081,
      "toolCalls": 4,
      "inputTokens": 751,
      "outputTokens": 318,
      "reasoningTokens": 0,
      "cacheReadTokens": 3259,
      "cacheWriteTokens": 157
    },
    {
      "trialId": "extension-v1:scored:sup-002:zai/glm-5.3-flash:1",
      "fixtureId": "sup-002",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0006850599999999999,
      "latencyMs": 8268,
      "toolCalls": 4,
      "inputTokens": 3338,
      "outputTokens": 306,
      "reasoningTokens": 32,
      "cacheReadTokens": 512,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-006:openai/gpt-6-astra:1",
      "fixtureId": "rec-006",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.048729999999999996,
      "latencyMs": 13293,
      "toolCalls": 4,
      "inputTokens": 3523,
      "outputTokens": 270,
      "reasoningTokens": 0,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-002:openai/gpt-6-astra:1",
      "fixtureId": "sup-002",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.028421000000000002,
      "latencyMs": 13295,
      "toolCalls": 4,
      "inputTokens": 803,
      "outputTokens": 300,
      "reasoningTokens": 0,
      "cacheReadTokens": 3341,
      "cacheWriteTokens": 164
    },
    {
      "trialId": "extension-v1:scored:sup-001:zai/glm-5.3-flash:1",
      "fixtureId": "sup-001",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00082409,
      "latencyMs": 12735,
      "toolCalls": 5,
      "inputTokens": 3337,
      "outputTokens": 540,
      "reasoningTokens": 61,
      "cacheReadTokens": 768,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-006:zai/glm-5.3-flash:2",
      "fixtureId": "rec-006",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00056689,
      "latencyMs": 7255,
      "toolCalls": 4,
      "inputTokens": 2479,
      "outputTokens": 303,
      "reasoningTokens": 41,
      "cacheReadTokens": 768,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-009:zai/glm-5.3-flash:1",
      "fixtureId": "rec-009",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0007846400000000001,
      "latencyMs": 10843,
      "toolCalls": 4,
      "inputTokens": 3254,
      "outputTokens": 400,
      "reasoningTokens": 147,
      "cacheReadTokens": 768,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-001:openai/gpt-6-astra:1",
      "fixtureId": "sup-001",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.030172000000000004,
      "latencyMs": 13342,
      "toolCalls": 5,
      "inputTokens": 750,
      "outputTokens": 341,
      "reasoningTokens": 0,
      "cacheReadTokens": 3497,
      "cacheWriteTokens": 170
    },
    {
      "trialId": "extension-v1:scored:rec-009:openai/gpt-6-astra:1",
      "fixtureId": "rec-009",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 1,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.029387000000000003,
      "latencyMs": 13339,
      "toolCalls": 4,
      "inputTokens": 885,
      "outputTokens": 340,
      "reasoningTokens": 0,
      "cacheReadTokens": 3537,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-006:openai/gpt-6-astra:2",
      "fixtureId": "rec-006",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.04843,
      "latencyMs": 13334,
      "toolCalls": 4,
      "inputTokens": 3523,
      "outputTokens": 264,
      "reasoningTokens": 0,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-005:openai/gpt-6-astra:2",
      "fixtureId": "ext-005",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.047409999999999994,
      "latencyMs": 10848,
      "toolCalls": 4,
      "inputTokens": 3516,
      "outputTokens": 245,
      "reasoningTokens": 0,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-005:zai/glm-5.3-flash:2",
      "fixtureId": "ext-005",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0006219800000000001,
      "latencyMs": 14125,
      "toolCalls": 4,
      "inputTokens": 2992,
      "outputTokens": 277,
      "reasoningTokens": 54,
      "cacheReadTokens": 256,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-003:openai/gpt-6-astra:2",
      "fixtureId": "rec-003",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.027999,
      "latencyMs": 14307,
      "toolCalls": 5,
      "inputTokens": 753,
      "outputTokens": 340,
      "reasoningTokens": 0,
      "cacheReadTokens": 3469,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-003:zai/glm-5.3-flash:2",
      "fixtureId": "sup-003",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00068863,
      "latencyMs": 10548,
      "toolCalls": 4,
      "inputTokens": 2667,
      "outputTokens": 435,
      "reasoningTokens": 50,
      "cacheReadTokens": 1536,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-009:openai/gpt-6-astra:2",
      "fixtureId": "sup-009",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0376765,
      "latencyMs": 16037,
      "toolCalls": 6,
      "inputTokens": 892,
      "outputTokens": 429,
      "reasoningTokens": 0,
      "cacheReadTokens": 4369,
      "cacheWriteTokens": 235
    },
    {
      "trialId": "extension-v1:scored:rec-003:zai/glm-5.3-flash:2",
      "fixtureId": "rec-003",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00087561,
      "latencyMs": 11796,
      "toolCalls": 7,
      "inputTokens": 4485,
      "outputTokens": 375,
      "reasoningTokens": 0,
      "cacheReadTokens": 512,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-003:openai/gpt-6-astra:2",
      "fixtureId": "sup-003",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.023725499999999997,
      "latencyMs": 13066,
      "toolCalls": 4,
      "inputTokens": 843,
      "outputTokens": 227,
      "reasoningTokens": 0,
      "cacheReadTokens": 2258,
      "cacheWriteTokens": 135
    },
    {
      "trialId": "extension-v1:scored:sup-009:zai/glm-5.3-flash:2",
      "fixtureId": "sup-009",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00102399,
      "latencyMs": 23349,
      "toolCalls": 6,
      "inputTokens": 4153,
      "outputTokens": 695,
      "reasoningTokens": 61,
      "cacheReadTokens": 768,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-007:zai/glm-5.3-flash:2",
      "fixtureId": "rec-007",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00068596,
      "latencyMs": 8350,
      "toolCalls": 6,
      "inputTokens": 3068,
      "outputTokens": 323,
      "reasoningTokens": 21,
      "cacheReadTokens": 1792,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-003:zai/glm-5.3-flash:2",
      "fixtureId": "ext-003",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0006201900000000001,
      "latencyMs": 8975,
      "toolCalls": 2,
      "inputTokens": 3181,
      "outputTokens": 230,
      "reasoningTokens": 10,
      "cacheReadTokens": 768,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-001:zai/glm-5.3-flash:2",
      "fixtureId": "rec-001",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00047304,
      "latencyMs": 8997,
      "toolCalls": 3,
      "inputTokens": 2250,
      "outputTokens": 200,
      "reasoningTokens": 25,
      "cacheReadTokens": 768,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-001:openai/gpt-6-astra:2",
      "fixtureId": "rec-001",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.042960000000000005,
      "latencyMs": 10225,
      "toolCalls": 3,
      "inputTokens": 3196,
      "outputTokens": 220,
      "reasoningTokens": 0,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-007:openai/gpt-6-astra:2",
      "fixtureId": "rec-007",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0503225,
      "latencyMs": 12000,
      "toolCalls": 5,
      "inputTokens": 1734,
      "outputTokens": 346,
      "reasoningTokens": 0,
      "cacheReadTokens": 1145,
      "cacheWriteTokens": 1163
    },
    {
      "trialId": "extension-v1:scored:ext-003:openai/gpt-6-astra:2",
      "fixtureId": "ext-003",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.024922,
      "latencyMs": 12904,
      "toolCalls": 2,
      "inputTokens": 951,
      "outputTokens": 239,
      "reasoningTokens": 0,
      "cacheReadTokens": 3462,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-004:openai/gpt-6-astra:2",
      "fixtureId": "rec-004",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.027768,
      "latencyMs": 12338,
      "toolCalls": 4,
      "inputTokens": 782,
      "outputTokens": 331,
      "reasoningTokens": 0,
      "cacheReadTokens": 3398,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-010:zai/glm-5.3-flash:2",
      "fixtureId": "ext-010",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0008288900000000001,
      "latencyMs": 6333,
      "toolCalls": 4,
      "inputTokens": 3909,
      "outputTokens": 295,
      "reasoningTokens": 144,
      "cacheReadTokens": 768,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-009:zai/glm-5.3-flash:2",
      "fixtureId": "ext-009",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00054776,
      "latencyMs": 6895,
      "toolCalls": 2,
      "inputTokens": 2626,
      "outputTokens": 269,
      "reasoningTokens": 8,
      "cacheReadTokens": 512,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-004:zai/glm-5.3-flash:2",
      "fixtureId": "rec-004",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00061535,
      "latencyMs": 16898,
      "toolCalls": 4,
      "inputTokens": 2453,
      "outputTokens": 397,
      "reasoningTokens": 21,
      "cacheReadTokens": 1280,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-008:zai/glm-5.3-flash:2",
      "fixtureId": "rec-008",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0006423799999999999,
      "latencyMs": 8482,
      "toolCalls": 4,
      "inputTokens": 3188,
      "outputTokens": 274,
      "reasoningTokens": 39,
      "cacheReadTokens": 256,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-008:openai/gpt-6-astra:2",
      "fixtureId": "rec-008",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.032964,
      "latencyMs": 10860,
      "toolCalls": 4,
      "inputTokens": 1652,
      "outputTokens": 286,
      "reasoningTokens": 0,
      "cacheReadTokens": 2144,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-009:openai/gpt-6-astra:2",
      "fixtureId": "ext-009",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0465,
      "latencyMs": 11387,
      "toolCalls": 2,
      "inputTokens": 3295,
      "outputTokens": 271,
      "reasoningTokens": 0,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-010:openai/gpt-6-astra:2",
      "fixtureId": "ext-010",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.04182,
      "latencyMs": 14724,
      "toolCalls": 4,
      "inputTokens": 2461,
      "outputTokens": 302,
      "reasoningTokens": 0,
      "cacheReadTokens": 2110,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-001:zai/glm-5.3-flash:2",
      "fixtureId": "ext-001",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00056248,
      "latencyMs": 6309,
      "toolCalls": 2,
      "inputTokens": 2506,
      "outputTokens": 281,
      "reasoningTokens": 0,
      "cacheReadTokens": 1536,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-001:openai/gpt-6-astra:2",
      "fixtureId": "ext-001",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.029301,
      "latencyMs": 12212,
      "toolCalls": 2,
      "inputTokens": 883,
      "outputTokens": 336,
      "reasoningTokens": 0,
      "cacheReadTokens": 3671,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-008:openai/gpt-6-astra:2",
      "fixtureId": "ext-008",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0522,
      "latencyMs": 18157,
      "toolCalls": 3,
      "inputTokens": 4025,
      "outputTokens": 239,
      "reasoningTokens": 0,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-008:zai/glm-5.3-flash:2",
      "fixtureId": "sup-008",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00122964,
      "latencyMs": 21429,
      "toolCalls": 7,
      "inputTokens": 4904,
      "outputTokens": 804,
      "reasoningTokens": 138,
      "cacheReadTokens": 768,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-008:openai/gpt-6-astra:2",
      "fixtureId": "sup-008",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.081931,
      "latencyMs": 21428,
      "toolCalls": 7,
      "inputTokens": 12,
      "outputTokens": 609,
      "reasoningTokens": 0,
      "cacheReadTokens": 2661,
      "cacheWriteTokens": 3896
    },
    {
      "trialId": "extension-v1:scored:ext-007:zai/glm-5.3-flash:2",
      "fixtureId": "ext-007",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00062499,
      "latencyMs": 12924,
      "toolCalls": 4,
      "inputTokens": 2953,
      "outputTokens": 284,
      "reasoningTokens": 34,
      "cacheReadTokens": 768,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-004:zai/glm-5.3-flash:2",
      "fixtureId": "sup-004",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0007954099999999999,
      "latencyMs": 14266,
      "toolCalls": 5,
      "inputTokens": 3757,
      "outputTokens": 378,
      "reasoningTokens": 55,
      "cacheReadTokens": 512,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-008:zai/glm-5.3-flash:2",
      "fixtureId": "ext-008",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00060673,
      "latencyMs": 14886,
      "toolCalls": 3,
      "inputTokens": 2671,
      "outputTokens": 250,
      "reasoningTokens": 70,
      "cacheReadTokens": 1536,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-004:openai/gpt-6-astra:2",
      "fixtureId": "sup-004",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.029994999999999997,
      "latencyMs": 17872,
      "toolCalls": 5,
      "inputTokens": 844,
      "outputTokens": 350,
      "reasoningTokens": 0,
      "cacheReadTokens": 4055,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-005:zai/glm-5.3-flash:2",
      "fixtureId": "rec-005",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00044819,
      "latencyMs": 3617,
      "toolCalls": 3,
      "inputTokens": 2111,
      "outputTokens": 178,
      "reasoningTokens": 39,
      "cacheReadTokens": 768,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-001:zai/glm-5.3-flash:2",
      "fixtureId": "sup-001",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00077184,
      "latencyMs": 9384,
      "toolCalls": 5,
      "inputTokens": 3262,
      "outputTokens": 449,
      "reasoningTokens": 70,
      "cacheReadTokens": 768,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-005:openai/gpt-6-astra:2",
      "fixtureId": "rec-005",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.039200000000000006,
      "latencyMs": 12612,
      "toolCalls": 3,
      "inputTokens": 2975,
      "outputTokens": 189,
      "reasoningTokens": 0,
      "cacheReadTokens": 0,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-007:openai/gpt-6-astra:2",
      "fixtureId": "ext-007",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.02639,
      "latencyMs": 12639,
      "toolCalls": 4,
      "inputTokens": 837,
      "outputTokens": 293,
      "reasoningTokens": 0,
      "cacheReadTokens": 3370,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-004:openai/gpt-6-astra:2",
      "fixtureId": "ext-004",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.030036,
      "latencyMs": 13338,
      "toolCalls": 6,
      "inputTokens": 745,
      "outputTokens": 383,
      "reasoningTokens": 0,
      "cacheReadTokens": 3436,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-001:openai/gpt-6-astra:2",
      "fixtureId": "sup-001",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.030253000000000002,
      "latencyMs": 13380,
      "toolCalls": 5,
      "inputTokens": 750,
      "outputTokens": 341,
      "reasoningTokens": 0,
      "cacheReadTokens": 3503,
      "cacheWriteTokens": 176
    },
    {
      "trialId": "extension-v1:scored:ext-004:zai/glm-5.3-flash:2",
      "fixtureId": "ext-004",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0006817,
      "latencyMs": 14136,
      "toolCalls": 6,
      "inputTokens": 2562,
      "outputTokens": 435,
      "reasoningTokens": 83,
      "cacheReadTokens": 1280,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-009:zai/glm-5.3-flash:2",
      "fixtureId": "rec-009",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00074709,
      "latencyMs": 7374,
      "toolCalls": 4,
      "inputTokens": 3217,
      "outputTokens": 373,
      "reasoningTokens": 110,
      "cacheReadTokens": 768,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-006:zai/glm-5.3-flash:2",
      "fixtureId": "sup-006",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00075384,
      "latencyMs": 8821,
      "toolCalls": 5,
      "inputTokens": 3432,
      "outputTokens": 392,
      "reasoningTokens": 40,
      "cacheReadTokens": 768,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-010:openai/gpt-6-astra:2",
      "fixtureId": "sup-010",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.029567,
      "latencyMs": 12993,
      "toolCalls": 4,
      "inputTokens": 904,
      "outputTokens": 329,
      "reasoningTokens": 0,
      "cacheReadTokens": 4077,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-006:openai/gpt-6-astra:2",
      "fixtureId": "ext-006",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.026776,
      "latencyMs": 12991,
      "toolCalls": 4,
      "inputTokens": 751,
      "outputTokens": 317,
      "reasoningTokens": 0,
      "cacheReadTokens": 3416,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-009:openai/gpt-6-astra:2",
      "fixtureId": "rec-009",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.029137000000000003,
      "latencyMs": 13137,
      "toolCalls": 4,
      "inputTokens": 885,
      "outputTokens": 335,
      "reasoningTokens": 0,
      "cacheReadTokens": 3537,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-006:zai/glm-5.3-flash:2",
      "fixtureId": "ext-006",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00065334,
      "latencyMs": 17351,
      "toolCalls": 4,
      "inputTokens": 2922,
      "outputTokens": 349,
      "reasoningTokens": 35,
      "cacheReadTokens": 768,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-010:zai/glm-5.3-flash:2",
      "fixtureId": "sup-010",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00073287,
      "latencyMs": 18939,
      "toolCalls": 4,
      "inputTokens": 3281,
      "outputTokens": 390,
      "reasoningTokens": 30,
      "cacheReadTokens": 1024,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-002:zai/glm-5.3-flash:2",
      "fixtureId": "rec-002",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00066755,
      "latencyMs": 7576,
      "toolCalls": 5,
      "inputTokens": 3261,
      "outputTokens": 259,
      "reasoningTokens": 21,
      "cacheReadTokens": 1280,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-006:openai/gpt-6-astra:2",
      "fixtureId": "sup-006",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.024565,
      "latencyMs": 14256,
      "toolCalls": 5,
      "inputTokens": 836,
      "outputTokens": 248,
      "reasoningTokens": 0,
      "cacheReadTokens": 3805,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:ext-002:zai/glm-5.3-flash:2",
      "fixtureId": "ext-002",
      "model": "zai/glm-5.3-flash",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00056473,
      "latencyMs": 5404,
      "toolCalls": 3,
      "inputTokens": 2251,
      "outputTokens": 306,
      "reasoningTokens": 56,
      "cacheReadTokens": 1536,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-002:openai/gpt-6-astra:2",
      "fixtureId": "rec-002",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.040754,
      "latencyMs": 11290,
      "toolCalls": 4,
      "inputTokens": 2607,
      "outputTokens": 273,
      "reasoningTokens": 0,
      "cacheReadTokens": 1034,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-010:openai/gpt-6-astra:2",
      "fixtureId": "rec-010",
      "model": "openai/gpt-6-astra",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.048674499999999996,
      "latencyMs": 13316,
      "toolCalls": 7,
      "inputTokens": 1651,
      "outputTokens": 310,
      "reasoningTokens": 0,
      "cacheReadTokens": 2177,
      "cacheWriteTokens": 1159
    },
    {
      "trialId": "extension-v1:scored:ext-002:openai/gpt-6-astra:2",
      "fixtureId": "ext-002",
      "model": "openai/gpt-6-astra",
      "workflow": "extraction",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.020368,
      "latencyMs": 7837,
      "toolCalls": 3,
      "inputTokens": 782,
      "outputTokens": 206,
      "reasoningTokens": 0,
      "cacheReadTokens": 2248,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-002:zai/glm-5.3-flash:2",
      "fixtureId": "sup-002",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00069486,
      "latencyMs": 4922,
      "toolCalls": 4,
      "inputTokens": 3350,
      "outputTokens": 317,
      "reasoningTokens": 37,
      "cacheReadTokens": 512,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-002:openai/gpt-6-astra:2",
      "fixtureId": "sup-002",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.027743999999999998,
      "latencyMs": 12908,
      "toolCalls": 4,
      "inputTokens": 803,
      "outputTokens": 287,
      "reasoningTokens": 0,
      "cacheReadTokens": 3339,
      "cacheWriteTokens": 162
    },
    {
      "trialId": "extension-v1:scored:sup-007:openai/gpt-6-astra:2",
      "fixtureId": "sup-007",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.029614,
      "latencyMs": 13653,
      "toolCalls": 7,
      "inputTokens": 798,
      "outputTokens": 325,
      "reasoningTokens": 0,
      "cacheReadTokens": 5384,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-005:openai/gpt-6-astra:2",
      "fixtureId": "sup-005",
      "model": "openai/gpt-6-astra",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0358545,
      "latencyMs": 14740,
      "toolCalls": 6,
      "inputTokens": 962,
      "outputTokens": 364,
      "reasoningTokens": 0,
      "cacheReadTokens": 5597,
      "cacheWriteTokens": 195
    },
    {
      "trialId": "extension-v1:scored:sup-005:zai/glm-5.3-flash:2",
      "fixtureId": "sup-005",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.0012563700000000001,
      "latencyMs": 16137,
      "toolCalls": 7,
      "inputTokens": 5721,
      "outputTokens": 480,
      "reasoningTokens": 255,
      "cacheReadTokens": 1024,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:rec-010:zai/glm-5.3-flash:2",
      "fixtureId": "rec-010",
      "model": "zai/glm-5.3-flash",
      "workflow": "reconciliation",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00066468,
      "latencyMs": 30864,
      "toolCalls": 8,
      "inputTokens": 3134,
      "outputTokens": 297,
      "reasoningTokens": 0,
      "cacheReadTokens": 1536,
      "cacheWriteTokens": 0
    },
    {
      "trialId": "extension-v1:scored:sup-007:zai/glm-5.3-flash:2",
      "fixtureId": "sup-007",
      "model": "zai/glm-5.3-flash",
      "workflow": "support",
      "trial": 2,
      "status": "completed",
      "passed": true,
      "reasons": [],
      "cost": 0.00104096,
      "latencyMs": 22165,
      "toolCalls": 8,
      "inputTokens": 5304,
      "outputTokens": 415,
      "reasoningTokens": 45,
      "cacheReadTokens": 512,
      "cacheWriteTokens": 0
    }
  ],
  "sourceHashes": {
    "originalReport": "0a06bb1b4f502ef11439aaa425a7d8b5a25094e8ba361a6aa05f8831b68f1041",
    "extensionReport": "b418ee8d1cd9368c737cd9e042079fc01752cb5dd73e0f16d5b17c6185204c74",
    "originalJournal": "5dfb708644d4c4a5e226f5c13b12e99960d8c52d3be445d6327c79b0f3082c92",
    "extensionJournal": "c37819de84431dc24ea97d2ed9159fc02b68f32e63148d3b6217b8bcc7b86217"
  }
}
