{
  "dataset": "BU_Bench_v2",
  "tasksPerSetting": 60,
  "datasetSha256": "fedc07de8d02a73e81fe1aa50775d753d00b7714203b87bbdd42e75a7257ffd0",
  "sourceUrl": "https://github.com/browser-use/benchmark/tree/1e0e2f1ed12d3dfbdfab6b7bff415ec090ff9f74#benchmark-scores--historical-60-task-results",
  "dataSourceUrl": "/benchmarks/pricing-models.json",
  "scoreDefinition": "Continuous findings quality score from 0 to 100, not a binary success rate.",
  "costDefinition": "Recorded agent cost per task, excluding judge, browser, runner, and unrecorded auxiliary calls.",
  "selection": "All models in the Cloud model registry with comparable recorded results in this snapshot; includes all recorded settings except Luna low with incomplete cost.",
  "rows": [
    {
      "id": "claude-opus-5-default",
      "label": "Claude Opus 5",
      "model": "Claude Opus 5",
      "effort": null,
      "score": 50.53333333333333,
      "cost": 7.370673837499999,
      "duration": 35.89768218199412,
      "n": 60,
      "cost_is_lower_bound": null,
      "caveat": null
    },
    {
      "id": "claude-opus-5-5-default",
      "label": "Claude Opus 5.5",
      "model": "Claude Opus 5.5",
      "effort": null,
      "score": 59.4,
      "cost": 4.214918616666667,
      "duration": 34.19066027777777,
      "n": 60,
      "cost_is_lower_bound": true,
      "caveat": null
    },
    {
      "id": "deepseek-v4.1-flash-high",
      "label": "DeepSeek V4.1 Flash · high",
      "model": "DeepSeek V4.1 Flash",
      "effort": "high",
      "score": 41.46666666666667,
      "cost": 0.15239166666666668,
      "duration": 16.265566666666665,
      "n": 60,
      "cost_is_lower_bound": true,
      "caveat": "completed main-loop calls; excludes judge/browser/runner and unreported auxiliary usage"
    },
    {
      "id": "deepseek-v4.1-flash-low",
      "label": "DeepSeek V4.1 Flash · low",
      "model": "DeepSeek V4.1 Flash",
      "effort": "low",
      "score": 38.166666666666664,
      "cost": 0.13841965,
      "duration": 16.978566666666666,
      "n": 60,
      "cost_is_lower_bound": true,
      "caveat": "Recorded main-loop agent calls from retained attempts. Excludes unmetered auxiliary/interrupted calls, cancelled attempt spend, judge, browser and runner. Not invoice spend."
    },
    {
      "id": "deepseek-v4.1-flash-max",
      "label": "DeepSeek V4.1 Flash · max",
      "model": "DeepSeek V4.1 Flash",
      "effort": "max",
      "score": 47.88333333333333,
      "cost": 0.26276373333333336,
      "duration": 31.814308333333333,
      "n": 60,
      "cost_is_lower_bound": true,
      "caveat": "Recorded main-loop agent calls from retained attempts. Excludes unmetered auxiliary/interrupted calls, cancelled attempt spend, judge, browser and runner. Not invoice spend."
    },
    {
      "id": "deepseek-v4.1-flash-none",
      "label": "DeepSeek V4.1 Flash · none",
      "model": "DeepSeek V4.1 Flash",
      "effort": "none",
      "score": 29.049999999999997,
      "cost": 0.11757063333333331,
      "duration": 17.38586666666667,
      "n": 60,
      "cost_is_lower_bound": true,
      "caveat": "Recorded main-loop agent calls from retained attempts. Excludes unmetered auxiliary/interrupted calls, cancelled attempt spend, judge, browser and runner. Not invoice spend."
    },
    {
      "id": "glm-5.3-flash-default",
      "label": "GLM 5.3 Flash",
      "model": "GLM 5.3 Flash",
      "effort": null,
      "score": 28.03333333333333,
      "cost": 0.11415670733333333,
      "duration": 28.762488238016765,
      "n": 60,
      "cost_is_lower_bound": null,
      "caveat": null
    },
    {
      "id": "gpt-5.6-luna-default",
      "label": "GPT-5.6 Luna · default",
      "model": "GPT-5.6 Luna",
      "effort": "default",
      "score": 16.05,
      "cost": 0.055631897333333326,
      "duration": 5.339573774735133,
      "n": 60,
      "cost_is_lower_bound": null,
      "caveat": null
    },
    {
      "id": "gpt-5.6-luna-xhigh",
      "label": "GPT-5.6 Luna · xhigh",
      "model": "GPT-5.6 Luna",
      "effort": "xhigh",
      "score": 41.16666666666667,
      "cost": 0.35812028433333337,
      "duration": 22.150595784187317,
      "n": 60,
      "cost_is_lower_bound": null,
      "caveat": null
    },
    {
      "id": "gpt-5.6-sol-default",
      "label": "GPT-5.6 Sol · default",
      "model": "GPT-5.6 Sol",
      "effort": "default",
      "score": 38.45,
      "cost": 2.7146040291666664,
      "duration": 13.230841330687205,
      "n": 60,
      "cost_is_lower_bound": null,
      "caveat": null
    },
    {
      "id": "gpt-5.6-sol-xhigh",
      "label": "GPT-5.6 Sol · xhigh",
      "model": "GPT-5.6 Sol",
      "effort": "xhigh",
      "score": 49.06666666666667,
      "cost": 7.803157195833334,
      "duration": 23.630619355042775,
      "n": 60,
      "cost_is_lower_bound": null,
      "caveat": null
    },
    {
      "id": "gpt-6-astra-medium",
      "label": "GPT-6 Astra · medium",
      "model": "GPT-6 Astra",
      "effort": "medium",
      "score": 80.63333333333334,
      "cost": 8.867206283333333,
      "duration": 31.674516666666666,
      "n": 60,
      "cost_is_lower_bound": true,
      "caveat": "Retained main-loop cost; child-agent and other auxiliary usage may be missing"
    },
    {
      "id": "gpt-6-luna-high",
      "label": "GPT-6 Luna · high",
      "model": "GPT-6 Luna",
      "effort": "high",
      "score": 53.949999999999996,
      "cost": 0.1547484,
      "duration": 25.921848055555554,
      "n": 60,
      "cost_is_lower_bound": true,
      "caveat": null
    },
    {
      "id": "gpt-6-luna-medium",
      "label": "GPT-6 Luna · medium",
      "model": "GPT-6 Luna",
      "effort": "medium",
      "score": 50.3,
      "cost": 0.12648466666666666,
      "duration": 19.809375,
      "n": 60,
      "cost_is_lower_bound": true,
      "caveat": null
    },
    {
      "id": "gpt-6-luna-none",
      "label": "GPT-6 Luna · none",
      "model": "GPT-6 Luna",
      "effort": "none",
      "score": 31.75,
      "cost": 0.07684865,
      "duration": 8.715164166666666,
      "n": 60,
      "cost_is_lower_bound": true,
      "caveat": null
    },
    {
      "id": "gpt-6-luna-xhigh",
      "label": "GPT-6 Luna · xhigh",
      "model": "GPT-6 Luna",
      "effort": "xhigh",
      "score": 57.550000000000004,
      "cost": 0.18953225,
      "duration": 30.740511388888887,
      "n": 60,
      "cost_is_lower_bound": true,
      "caveat": null
    },
    {
      "id": "gpt-6-sol-medium",
      "label": "GPT-6 Sol · medium",
      "model": "GPT-6 Sol",
      "effort": "medium",
      "score": 66.85,
      "cost": 1.2027994166666667,
      "duration": 13.527046666666667,
      "n": 60,
      "cost_is_lower_bound": true,
      "caveat": null
    },
    {
      "id": "grok-4.5-default",
      "label": "Grok 4.5",
      "model": "Grok 4.5",
      "effort": null,
      "score": 33.03333333333333,
      "cost": 1.15309032,
      "duration": 16.345784443616868,
      "n": 60,
      "cost_is_lower_bound": null,
      "caveat": null
    },
    {
      "id": "grok-4.6-default",
      "label": "Grok 4.6",
      "model": "Grok 4.6",
      "effort": null,
      "score": 31.183333333333334,
      "cost": 3.934024266666667,
      "duration": 23.3875592648983,
      "n": 60,
      "cost_is_lower_bound": null,
      "caveat": null
    },
    {
      "id": "minimax-m3-default",
      "label": "MiniMax M3",
      "model": "MiniMax M3",
      "effort": null,
      "score": 11.816666666666666,
      "cost": 0.883116569,
      "duration": 28.243562364578246,
      "n": 60,
      "cost_is_lower_bound": null,
      "caveat": null
    }
  ],
  "snapshotCommit": "b34e067bb380dc197cfd9a202e09d3cd2dd9b6fb",
  "cloudModelSource": "browser-use/cloud@d8e95bfc656f028866da34052f70714f8d97c7ab:frontend/src/lib/agent-models.ts",
  "withoutComparableResults": [
    "BU Ultrafast",
    "BU Fast",
    "GPT-6.1 Sol",
    "MiMo V2.6 Pro",
    "MiMo V2.6 Flash"
  ]
}