{
  "catalogVersion": "4510bd969166c4aba5fab0e6303d728328ea2ea027158d72b26b3ec40ce75586",
  "benchmarkVersion": "a33a08b026c25d98d18c7475afab565beb29574abc5a6793eff34f03c3176f5a",
  "cohorts": [
    {
      "id": "aa-intelligence-index-v5-2026-11-19",
      "seriesId": "aa-intelligence-index-v5",
      "title": "Artificial Analysis Intelligence Index",
      "version": "v5",
      "unit": "index points",
      "conditions": "Independent evaluation by Artificial Analysis on its standard harness, each model at the configuration named with its score.",
      "sourceId": "artificial-analysis",
      "measuredAt": "2026-11-19",
      "reportedAt": "2026-11-19",
      "verifiedAt": "2026-11-19",
      "observations": [
        {
          "modelId": "claude-fable-5-2",
          "value": 64.6,
          "configuration": "Claude Fable 5.2 (Max, Default Fallback)"
        },
        {
          "modelId": "gpt-6-2-astra",
          "value": 62.1,
          "configuration": "GPT-6.2 Astra (Max)"
        },
        {
          "modelId": "gemini-4-carbon",
          "value": 60,
          "configuration": "Gemini 4 Carbon (High)"
        },
        {
          "modelId": "claude-opus-5-5",
          "value": 59.1,
          "configuration": "Claude Opus 5.5 (Max, Default Fallback)"
        },
        {
          "modelId": "gpt-6-2-sol",
          "value": 57.3,
          "configuration": "GPT-6.2 Sol (Max)"
        },
        {
          "modelId": "claude-sonnet-5-5",
          "value": 54.5,
          "configuration": "Claude Sonnet 5.5 (Max, Default Fallback)"
        },
        {
          "modelId": "gemini-4-hydrogen",
          "value": 53.9,
          "configuration": "Gemini 4 Hydrogen (Xhigh)"
        },
        {
          "modelId": "gpt-6-1-terra",
          "value": 51.3,
          "configuration": "GPT-6.1 Terra (Max)"
        },
        {
          "modelId": "gemini-4-argon",
          "value": 50.4,
          "configuration": "Gemini 4 Argon (High)"
        },
        {
          "modelId": "claude-haiku-5-5",
          "value": 36.5,
          "configuration": "Claude Haiku 5.5 (Max)"
        },
        {
          "modelId": "grok-4-7",
          "value": 35.5,
          "configuration": "Grok 4.7 (Xhigh)"
        },
        {
          "modelId": "qwen3-8-max-0902",
          "value": 33.7,
          "configuration": "Qwen3.8 Max (0902)"
        },
        {
          "modelId": "deepseek-v4-1-flash",
          "value": 32.6,
          "configuration": "DeepSeek V4.1 Flash 1017 (Max)"
        },
        {
          "modelId": "gpt-5-6-terra",
          "value": 32.3,
          "configuration": "GPT-5.6 Terra (Max)"
        },
        {
          "modelId": "gemini-3-8-flash",
          "value": 30.8,
          "configuration": "Gemini 3.8 Flash (High)"
        },
        {
          "modelId": "gpt-6-luna",
          "value": 27.9,
          "configuration": "GPT-6 Luna (Max)"
        },
        {
          "modelId": "mistral-large-4",
          "value": 26.1,
          "configuration": "Mistral Large 4 (High)"
        },
        {
          "modelId": "deepseek-v4-pro",
          "value": 23.1,
          "configuration": "DeepSeek V4 Pro 0813 (Max)"
        },
        {
          "modelId": "qwen3-8-27b",
          "value": 20.1,
          "configuration": "Qwen3.8 27B (Xhigh)"
        }
      ]
    },
    {
      "title": "Terminal-Bench",
      "version": "4.1",
      "unit": "percentage points",
      "sourceId": "terminal-bench",
      "measuredAt": "2026-11-19",
      "reportedAt": "2026-11-19",
      "verifiedAt": "2026-11-19",
      "id": "terminal-bench-4-1-models-2026-11-19",
      "seriesId": "terminal-bench-4-1-models",
      "category": "models",
      "conditions": "Best public leaderboard resolution rate for each model in any agent harness. The agent and effort behind the score are named with it.",
      "observations": [
        {
          "productId": "claude-fable-5-2",
          "value": 65.76,
          "configuration": "Claude Code · max",
          "detail": "±2.9 (95% CI), 330 trials, submitted 2026-11-09"
        },
        {
          "productId": "claude-opus-5-5",
          "value": 65.76,
          "configuration": "Claude Code · max",
          "detail": "±3.1 (95% CI), 330 trials, submitted 2026-09-22"
        },
        {
          "productId": "gemini-4-carbon",
          "value": 64.55,
          "configuration": "Antigravity CLI · high",
          "detail": "±3.2 (95% CI), 330 trials, submitted 2026-10-28"
        },
        {
          "productId": "claude-sonnet-5-5",
          "value": 63.94,
          "configuration": "Claude Code · max",
          "detail": "±3.0 (95% CI), 330 trials, submitted 2026-09-28"
        },
        {
          "productId": "gpt-6-2-astra",
          "value": 62.12,
          "configuration": "Codex · max",
          "detail": "±2.3 (95% CI), 330 trials, submitted 2026-11-17"
        },
        {
          "productId": "gpt-6-2-sol",
          "value": 61.52,
          "configuration": "Codex · max",
          "detail": "±2.8 (95% CI), 330 trials, submitted 2026-11-17"
        },
        {
          "productId": "gemini-4-hydrogen",
          "value": 56.36,
          "configuration": "Antigravity CLI · high",
          "detail": "±3.0 (95% CI), 330 trials, submitted 2026-11-12"
        },
        {
          "productId": "gemini-4-argon",
          "value": 55.45,
          "configuration": "Antigravity CLI · high",
          "detail": "±2.9 (95% CI), 330 trials, submitted 2026-09-30"
        },
        {
          "productId": "gpt-6-1-terra",
          "value": 53.33,
          "configuration": "Codex · max",
          "detail": "±3.1 (95% CI), 330 trials, submitted 2026-10-26"
        },
        {
          "productId": "claude-haiku-5-5",
          "value": 40,
          "configuration": "Claude Code · max",
          "detail": "±3.6 (95% CI), 330 trials, submitted 2026-10-07"
        },
        {
          "productId": "grok-4-7",
          "value": 38.79,
          "configuration": "Grok Build · xhigh",
          "detail": "±3.6 (95% CI), 330 trials, submitted 2026-09-21"
        },
        {
          "productId": "deepseek-v4-1-flash",
          "value": 38.48,
          "configuration": "Claude Code · max",
          "detail": "±3.8 (95% CI), 330 trials, submitted 2026-10-17"
        },
        {
          "productId": "mistral-large-4",
          "value": 27.58,
          "configuration": "Claude Code · max",
          "detail": "±3.9 (95% CI), 330 trials, submitted 2026-10-06"
        },
        {
          "productId": "qwen3-8-max-0902",
          "value": 26.97,
          "configuration": "Claude Code · max",
          "detail": "±3.9 (95% CI), 330 trials, submitted 2026-09-02"
        },
        {
          "productId": "gpt-5-6-terra",
          "value": 21.52,
          "configuration": "Codex · max",
          "detail": "±3.3 (95% CI), 330 trials, submitted 2026-06-26"
        },
        {
          "productId": "gemini-3-8-flash",
          "value": 18.18,
          "configuration": "mini-SWE-agent · high",
          "detail": "±3.4 (95% CI), 330 trials, submitted 2026-09-02"
        },
        {
          "productId": "gpt-6-luna",
          "value": 15.15,
          "configuration": "Codex · max",
          "detail": "±2.7 (95% CI), 330 trials, submitted 2026-09-22"
        },
        {
          "productId": "deepseek-v4-pro",
          "value": 6.67,
          "configuration": "Claude Code · max",
          "detail": "±2.5 (95% CI), 330 trials, submitted 2026-08-13"
        }
      ]
    },
    {
      "title": "Terminal-Bench",
      "version": "4.1",
      "unit": "percentage points",
      "sourceId": "terminal-bench",
      "measuredAt": "2026-11-19",
      "reportedAt": "2026-11-19",
      "verifiedAt": "2026-11-19",
      "id": "terminal-bench-4-1-agents-2026-11-19",
      "seriesId": "terminal-bench-4-1-agents",
      "category": "agents",
      "conditions": "Best public leaderboard resolution rate for each agent with any model. The model behind the score is named with it, and it drives much of the result.",
      "observations": [
        {
          "productId": "github-copilot",
          "value": 71.21,
          "configuration": "Copilot Fusion · max",
          "detail": "±2.9 (95% CI), 330 trials, submitted 2026-11-17"
        },
        {
          "productId": "claude-code",
          "value": 65.76,
          "configuration": "Fable 5.2 · max",
          "detail": "±2.9 (95% CI), 330 trials, submitted 2026-11-09"
        },
        {
          "productId": "antigravity-cli",
          "value": 64.55,
          "configuration": "Gemini 4 Carbon · high",
          "detail": "±3.2 (95% CI), 330 trials, submitted 2026-10-28"
        },
        {
          "productId": "codex",
          "value": 62.12,
          "configuration": "GPT-6.2 Astra · max",
          "detail": "±2.3 (95% CI), 330 trials, submitted 2026-11-17"
        },
        {
          "productId": "muse-code",
          "value": 46.97,
          "configuration": "Muse Wonder · xhigh",
          "detail": "±3.8 (95% CI), 330 trials, submitted 2026-10-12"
        },
        {
          "productId": "grok-build",
          "value": 38.79,
          "configuration": "Grok 4.7 · xhigh",
          "detail": "±3.6 (95% CI), 330 trials, submitted 2026-09-21"
        },
        {
          "productId": "mini-swe-agent",
          "value": 26.97,
          "configuration": "MiMo V2.6 Pro · none",
          "detail": "±3.6 (95% CI), 330 trials, submitted 2026-09-22"
        }
      ]
    }
  ],
  "trends": []
}