{
  "url": "https://www.aibattle.space/rankings/eval-tau3-3-banking-score",
  "dataset": "𝜏³-Banking: Score",
  "family": "tab::eval-tau3",
  "familyLabel": "eval tau3",
  "sourceRowCount": 20,
  "window": 20,
  "metricKeys": [
    "𝜏³-Banking"
  ],
  "rebuilt": "2026-09-19",
  "stats": {
    "median": 0.472164948453608,
    "mean": 0.45958762886597954,
    "min": 0.381443298969072,
    "max": 0.51340206185567,
    "spread": 0.13195876288659808,
    "gapFirstToSecond": 0.006185567010309034,
    "gapFirstToSecondPercent": 1.22,
    "creators": 10
  },
  "entries": [
    {
      "rank": 1,
      "name": "Qwen3.8 Max",
      "creator": "Alibaba",
      "slug": "qwen3-8-max",
      "detailsUrl": "/models/qwen3-8-max",
      "value": 0.51340206185567
    },
    {
      "rank": 2,
      "name": "Grok 4.6 (high)",
      "creator": "SpaceXAI",
      "slug": "grok-4-6",
      "detailsUrl": "/models/grok-4-6",
      "value": 0.507216494845361
    },
    {
      "rank": 3,
      "name": "Muse Spark 1.3 (max)",
      "creator": "Meta",
      "slug": "muse-spark-1-3",
      "detailsUrl": "/models/muse-spark-1-3",
      "value": 0.505154639175258
    },
    {
      "rank": 4,
      "name": "GLM-5.3 (max)",
      "creator": "Z AI",
      "slug": "glm-5-3",
      "detailsUrl": "/models/glm-5-3",
      "value": 0.503092783505155
    },
    {
      "rank": 5,
      "name": "Qwen3.8 2.4T A95B",
      "creator": "Alibaba",
      "slug": "qwen3-8-2-4t-a95b",
      "detailsUrl": "/models/qwen3-8-2-4t-a95b",
      "value": 0.490721649484536
    },
    {
      "rank": 6,
      "name": "Qwen3.8 27B (xhigh)",
      "creator": "Alibaba",
      "slug": "qwen3-8-27b",
      "detailsUrl": "/models/qwen3-8-27b",
      "value": 0.480412371134021
    },
    {
      "rank": 7,
      "name": "Agnes 3.0 Flash",
      "creator": "Sapiens AI",
      "slug": "agnes-3-0-flash",
      "detailsUrl": "/models/agnes-3-0-flash",
      "value": 0.476288659793814
    },
    {
      "rank": 8,
      "name": "Qwen3.8 27B (medium)",
      "creator": "Alibaba",
      "slug": "qwen3-8-27b-medium",
      "detailsUrl": "/models/qwen3-8-27b-medium",
      "value": 0.474226804123711
    },
    {
      "rank": 9,
      "name": "Claude Fable 5.1 (Adaptive Reasoning, Max Effort, Default Fallback)",
      "creator": "Anthropic",
      "slug": "claude-fable-5-1",
      "detailsUrl": "/models/claude-fable-5-1",
      "value": 0.472164948453608
    },
    {
      "rank": 10,
      "name": "Muse Spark 1.3 (xhigh)",
      "creator": "Meta",
      "slug": "muse-spark-1-3-xhigh",
      "detailsUrl": "/models/muse-spark-1-3-xhigh",
      "value": 0.472164948453608
    },
    {
      "rank": 11,
      "name": "GLM-5.3-Flash",
      "creator": "Z AI",
      "slug": "glm-5-3-flash",
      "detailsUrl": "/models/glm-5-3-flash",
      "value": 0.472164948453608
    },
    {
      "rank": 12,
      "name": "Kimi K3 (max)",
      "creator": "Kimi",
      "slug": "kimi-k3",
      "detailsUrl": "/models/kimi-k3",
      "value": 0.45979381443299
    },
    {
      "rank": 13,
      "name": "Claude Fable 5.1 (Adaptive Reasoning, Xhigh Effort, Default Fallback)",
      "creator": "Anthropic",
      "slug": "claude-fable-5-1-xhigh",
      "detailsUrl": "/models/claude-fable-5-1-xhigh",
      "value": 0.457731958762887
    },
    {
      "rank": 14,
      "name": "Gemini 3.8 Flash (high)",
      "creator": "Google",
      "slug": "gemini-3-8-flash",
      "detailsUrl": "/models/gemini-3-8-flash",
      "value": 0.449484536082474
    },
    {
      "rank": 15,
      "name": "GPT-5.6 Sol (max)",
      "creator": "OpenAI",
      "slug": "gpt-5-6-sol",
      "detailsUrl": "/models/gpt-5-6-sol",
      "value": 0.443298969072165
    },
    {
      "rank": 16,
      "name": "Claude Opus 5 (Adaptive Reasoning, Max Effort)",
      "creator": "Anthropic",
      "slug": "claude-opus-5",
      "detailsUrl": "/models/claude-opus-5",
      "value": 0.420618556701031
    },
    {
      "rank": 17,
      "name": "GPT-6 Astra (max)",
      "creator": "OpenAI",
      "slug": "gpt-6-astra",
      "detailsUrl": "/models/gpt-6-astra",
      "value": 0.414432989690722
    },
    {
      "rank": 18,
      "name": "GPT-5.6 Terra (max)",
      "creator": "OpenAI",
      "slug": "gpt-5-6-terra",
      "detailsUrl": "/models/gpt-5-6-terra",
      "value": 0.402061855670103
    },
    {
      "rank": 19,
      "name": "DeepSeek V4 Pro 0813 (Reasoning, Max Effort)",
      "creator": "DeepSeek",
      "slug": "deepseek-v4-pro",
      "detailsUrl": "/models/deepseek-v4-pro",
      "value": 0.395876288659794
    },
    {
      "rank": 20,
      "name": "Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)",
      "creator": "Anthropic",
      "slug": "claude-fable-5",
      "detailsUrl": "/models/claude-fable-5",
      "value": 0.381443298969072
    }
  ]
}