{
 "boards": [
  {
   "category": "general",
   "description": "Composite capability score fitted across many benchmarks.",
   "higher_is_better": true,
   "id": "epoch-eci",
   "metric": "ECI",
   "name": "Epoch Capabilities Index",
   "rows": [
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 178.79,
      "ci_low": 170.12
     },
     "model": "Claude Fable 5.2",
     "model_key": "anthropic/claude-fable-5-2",
     "provider": "anthropic",
     "rank": 1,
     "score": 174.24
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 172.51,
      "ci_low": 166.19
     },
     "model": "GPT-6 Astra",
     "model_key": "openai/gpt-6-astra",
     "provider": "openai",
     "rank": 2,
     "score": 169.16
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 172.24,
      "ci_low": 165.9
     },
     "model": "GPT-6.1 Sol",
     "model_key": "openai/gpt-6-1-sol",
     "provider": "openai",
     "rank": 3,
     "score": 168.87
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 171.91,
      "ci_low": 165.78
     },
     "model": "Claude Opus 5.5",
     "model_key": "anthropic/claude-opus-5-5",
     "provider": "anthropic",
     "rank": 4,
     "score": 168.67
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 171.5,
      "ci_low": 165.09
     },
     "model": "GPT-6.1 Terra",
     "model_key": "openai/gpt-6-1-terra",
     "provider": "openai",
     "rank": 5,
     "score": 168.11
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 170.2,
      "ci_low": 164.29
     },
     "model": "Claude Fable 5.1",
     "model_key": "anthropic/claude-fable-5-1",
     "provider": "anthropic",
     "rank": 6,
     "score": 167.04
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 169.53,
      "ci_low": 163.88
     },
     "model": "Gemini 4 Argon",
     "model_key": "google/gemini-4-argon",
     "provider": "google",
     "rank": 7,
     "score": 166.53
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 169.35,
      "ci_low": 163.65
     },
     "model": "Claude Sonnet 5.5",
     "model_key": "anthropic/claude-sonnet-5-5",
     "provider": "anthropic",
     "rank": 8,
     "score": 166.32
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 167.3,
      "ci_low": 162.02
     },
     "model": "Claude Opus 5",
     "model_key": "anthropic/claude-opus-5",
     "provider": "anthropic",
     "rank": 9,
     "score": 164.48
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 167.02,
      "ci_low": 161.83
     },
     "model": "Claude Fable 5",
     "model_key": "anthropic/claude-fable-5",
     "provider": "anthropic",
     "rank": 10,
     "score": 164.25
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 166.81,
      "ci_low": 161.51
     },
     "model": "GPT-6 Sol",
     "model_key": "openai/gpt-6-sol",
     "provider": "openai",
     "rank": 11,
     "score": 164
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 165.59,
      "ci_low": 160.72
     },
     "model": "GPT-5.6 Sol",
     "model_key": "openai/gpt-5-6-sol",
     "provider": "openai",
     "rank": 12,
     "score": 163.02
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 165.79,
      "ci_low": 159.99
     },
     "model": "GPT-5.5 Pro",
     "model_key": "openai/gpt-5-5-pro",
     "provider": "openai",
     "rank": 13,
     "score": 162.75
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 164.95,
      "ci_low": 159.45
     },
     "model": "Kimi K3.1 Pro",
     "model_key": "moonshotai/kimi-k3-1-pro",
     "provider": "moonshotai",
     "rank": 14,
     "score": 162.02
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 164.67,
      "ci_low": 159.6
     },
     "model": "Muse Wonder",
     "model_key": "meta-llama/muse-wonder",
     "provider": "meta-llama",
     "rank": 15,
     "score": 162.01
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 163.3,
      "ci_low": 158.62
     },
     "model": "GPT-5.6 Terra",
     "model_key": "openai/gpt-5-6-terra",
     "provider": "openai",
     "rank": 16,
     "score": 160.86
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 162.73,
      "ci_low": 158.02
     },
     "model": "Claude Haiku 5.5",
     "model_key": "anthropic/claude-haiku-5-5",
     "provider": "anthropic",
     "rank": 17,
     "score": 160.25
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 162.87,
      "ci_low": 157.72
     },
     "model": "GPT-5.4 Pro",
     "model_key": "openai/gpt-5-4-pro",
     "provider": "openai",
     "rank": 18,
     "score": 160.17
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 162.38,
      "ci_low": 158.02
     },
     "model": "GPT-5.5",
     "model_key": "openai/gpt-5-5",
     "provider": "openai",
     "rank": 19,
     "score": 160.06
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 162.25,
      "ci_low": 157.82
     },
     "model": "Claude Opus 4.8",
     "model_key": "anthropic/claude-opus-4-8",
     "provider": "anthropic",
     "rank": 20,
     "score": 159.91
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 161.86,
      "ci_low": 157.08
     },
     "model": "DeepSeek V4.1 Flash 1017",
     "model_key": "deepseek/deepseek-v4-1-flash-1017",
     "provider": "deepseek",
     "rank": 21,
     "score": 159.34
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 160.92,
      "ci_low": 156.59
     },
     "model": "Gemini 3.8 Flash",
     "model_key": "google/gemini-3-8-flash",
     "provider": "google",
     "rank": 22,
     "score": 158.66
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 160.71,
      "ci_low": 156.54
     },
     "model": "Kimi K3",
     "model_key": "moonshotai/kimi-k3",
     "provider": "moonshotai",
     "rank": 23,
     "score": 158.53
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 160.62,
      "ci_low": 156.44
     },
     "model": "Gemini 3.7 Flash",
     "model_key": "google/gemini-3-7-flash",
     "provider": "google",
     "rank": 24,
     "score": 158.44
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 160.26,
      "ci_low": 155.86
     },
     "model": "Muse Spark 1.3",
     "model_key": "meta-llama/muse-spark-1-3",
     "provider": "meta-llama",
     "rank": 25,
     "score": 157.97
    }
   ],
   "source": {
    "attribution": "Epoch AI",
    "licence": "CC-BY-4.0",
    "name": "Epoch AI",
    "url": "https://epoch.ai/eci",
    "via": null
   },
   "unit": "ECI",
   "updated_at": "2026-11-09"
  },
  {
   "category": "reasoning",
   "description": "Graduate-level multiple-choice science questions (Epoch's own runs).",
   "higher_is_better": true,
   "id": "gpqa-diamond",
   "metric": "accuracy",
   "name": "GPQA Diamond",
   "rows": [
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.1918
     },
     "model": "Gemini 4 Argon",
     "model_key": "google/gemini-4-argon",
     "provider": "google",
     "rank": 1,
     "score": 96.4015
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.3697
     },
     "model": "GPT-6 Astra",
     "model_key": "openai/gpt-6-astra",
     "provider": "openai",
     "rank": 2,
     "score": 95.7702
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.371
     },
     "model": "Claude Sonnet 5.5",
     "model_key": "anthropic/claude-sonnet-5-5",
     "provider": "anthropic",
     "rank": 3,
     "score": 95.5808
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.4011
     },
     "model": "Gemini 3.8 Flash",
     "model_key": "google/gemini-3-8-flash",
     "provider": "google",
     "rank": 4,
     "score": 95.3914
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.3781
     },
     "model": "GPT-6.1 Sol",
     "model_key": "openai/gpt-6-1-sol",
     "provider": "openai",
     "rank": 5,
     "score": 95.3914
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.3457
     },
     "model": "Gemini 3.7 Flash",
     "model_key": "google/gemini-3-7-flash",
     "provider": "google",
     "rank": 6,
     "score": 94.8232
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.6
     },
     "model": "GPT-5.4 Pro",
     "model_key": "openai/gpt-5-4-pro",
     "provider": "openai",
     "rank": 7,
     "score": 94.6
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.632
     },
     "model": "Gemini 3.1 Pro Preview",
     "model_key": "google/gemini-3-1-pro-preview",
     "provider": "google",
     "rank": 8,
     "score": 94.4444
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.4939
     },
     "model": "GPT-6 Sol",
     "model_key": "openai/gpt-6-sol",
     "provider": "openai",
     "rank": 9,
     "score": 94.3182
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.4912
     },
     "model": "GPT-6.1 Terra",
     "model_key": "openai/gpt-6-1-terra",
     "provider": "openai",
     "rank": 10,
     "score": 94.3182
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.4394
     },
     "model": "Kimi K3.1 Pro",
     "model_key": "moonshotai/kimi-k3-1-pro",
     "provider": "moonshotai",
     "rank": 11,
     "score": 94.1919
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.4029
     },
     "model": "Gemini 3.6 Flash",
     "model_key": "google/gemini-3-6-flash",
     "provider": "google",
     "rank": 12,
     "score": 94.1288
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.547
     },
     "model": "GPT-5.5",
     "model_key": "openai/gpt-5-5",
     "provider": "openai",
     "rank": 13,
     "score": 94.0025
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.448
     },
     "model": "Grok 4.6",
     "model_key": "x-ai/grok-4-6",
     "provider": "x-ai",
     "rank": 14,
     "score": 94.0025
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.5833
     },
     "model": "GPT-5.5 Pro",
     "model_key": "openai/gpt-5-5-pro",
     "provider": "openai",
     "rank": 15,
     "score": 93.9214
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.6253
     },
     "model": "Claude Fable 5.2",
     "model_key": "anthropic/claude-fable-5-2",
     "provider": "anthropic",
     "rank": 16,
     "score": 93.8763
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.4782
     },
     "model": "Claude Opus 5",
     "model_key": "anthropic/claude-opus-5",
     "provider": "anthropic",
     "rank": 17,
     "score": 93.8763
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.5701
     },
     "model": "GPT-5.6 Sol",
     "model_key": "openai/gpt-5-6-sol",
     "provider": "openai",
     "rank": 18,
     "score": 93.4975
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.434
     },
     "model": "Grok 4.5",
     "model_key": "x-ai/grok-4-5",
     "provider": "x-ai",
     "rank": 19,
     "score": 93.4343
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.5364
     },
     "model": "GPT-5.6 Terra",
     "model_key": "openai/gpt-5-6-terra",
     "provider": "openai",
     "rank": 20,
     "score": 93.3081
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.8
     },
     "model": "GPT-5.4",
     "model_key": "openai/gpt-5-4",
     "provider": "openai",
     "rank": 21,
     "score": 93.3
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.5671
     },
     "model": "DeepSeek V4.1 Flash 1017",
     "model_key": "deepseek/deepseek-v4-1-flash-1017",
     "provider": "deepseek",
     "rank": 22,
     "score": 93.1818
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.567
     },
     "model": "Doubao Seed Evolving",
     "model_key": "bytedance/doubao-seed-evolving",
     "provider": "bytedance",
     "rank": 23,
     "score": 93.1187
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.4937
     },
     "model": "Kimi K3",
     "model_key": "moonshotai/kimi-k3",
     "provider": "moonshotai",
     "rank": 24,
     "score": 93.1187
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.6387
     },
     "model": "Gemini 3.5 Flash",
     "model_key": "google/gemini-3-5-flash",
     "provider": "google",
     "rank": 25,
     "score": 92.803
    }
   ],
   "source": {
    "attribution": "Epoch AI",
    "licence": "CC-BY-4.0",
    "name": "Epoch AI",
    "url": "https://epoch.ai/benchmarks/gpqa-diamond",
    "via": null
   },
   "unit": "%",
   "updated_at": "2026-11-06"
  },
  {
   "category": "math",
   "description": "Unpublished research-level math problems (Epoch's own runs).",
   "higher_is_better": true,
   "id": "frontiermath",
   "metric": "accuracy",
   "name": "FrontierMath (Tiers 1-3)",
   "rows": [
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.2825
     },
     "model": "Claude Fable 5.2",
     "model_key": "anthropic/claude-fable-5-2",
     "provider": "anthropic",
     "rank": 1,
     "score": 95.0877
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.4434
     },
     "model": "GPT-6 Astra",
     "model_key": "openai/gpt-6-astra",
     "provider": "openai",
     "rank": 2,
     "score": 93.6842
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.4434
     },
     "model": "GPT-6.1 Sol",
     "model_key": "openai/gpt-6-1-sol",
     "provider": "openai",
     "rank": 3,
     "score": 93.6842
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.6786
     },
     "model": "Claude Opus 5.5",
     "model_key": "anthropic/claude-opus-5-5",
     "provider": "anthropic",
     "rank": 4,
     "score": 91.2281
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.7086
     },
     "model": "GPT-6.1 Terra",
     "model_key": "openai/gpt-6-1-terra",
     "provider": "openai",
     "rank": 5,
     "score": 90.8772
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.7662
     },
     "model": "Claude Fable 5.1",
     "model_key": "anthropic/claude-fable-5-1",
     "provider": "anthropic",
     "rank": 6,
     "score": 90.1754
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.794
     },
     "model": "GPT-6 Sol",
     "model_key": "openai/gpt-6-sol",
     "provider": "openai",
     "rank": 7,
     "score": 89.8246
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.8475
     },
     "model": "GPT-5.6 Sol",
     "model_key": "openai/gpt-5-6-sol",
     "provider": "openai",
     "rank": 8,
     "score": 89.1228
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.8734
     },
     "model": "Claude Sonnet 5.5",
     "model_key": "anthropic/claude-sonnet-5-5",
     "provider": "anthropic",
     "rank": 9,
     "score": 88.7719
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.9476
     },
     "model": "GPT-5.5 Pro",
     "model_key": "openai/gpt-5-5-pro",
     "provider": "openai",
     "rank": 10,
     "score": 87.7193
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.9944
     },
     "model": "Claude Fable 5",
     "model_key": "anthropic/claude-fable-5",
     "provider": "anthropic",
     "rank": 11,
     "score": 87.0175
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.0611
     },
     "model": "Gemini 4 Argon",
     "model_key": "google/gemini-4-argon",
     "provider": "google",
     "rank": 12,
     "score": 85.9649
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.0611
     },
     "model": "GPT-5.6 Terra",
     "model_key": "openai/gpt-5-6-terra",
     "provider": "openai",
     "rank": 13,
     "score": 85.9649
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.0825
     },
     "model": "Claude Opus 5",
     "model_key": "anthropic/claude-opus-5",
     "provider": "anthropic",
     "rank": 14,
     "score": 85.614
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.1034
     },
     "model": "GPT-5.5",
     "model_key": "openai/gpt-5-5",
     "provider": "openai",
     "rank": 15,
     "score": 85.2632
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.2569
     },
     "model": "GPT-5.4 Pro",
     "model_key": "openai/gpt-5-4-pro",
     "provider": "openai",
     "rank": 16,
     "score": 82.4561
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.2745
     },
     "model": "GPT-5.6 Luna",
     "model_key": "openai/gpt-5-6-luna",
     "provider": "openai",
     "rank": 17,
     "score": 82.1053
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.2745
     },
     "model": "Muse Wonder",
     "model_key": "meta-llama/muse-wonder",
     "provider": "meta-llama",
     "rank": 18,
     "score": 82.1053
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.3736
     },
     "model": "Claude Opus 4.8",
     "model_key": "anthropic/claude-opus-4-8",
     "provider": "anthropic",
     "rank": 19,
     "score": 80
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.4192
     },
     "model": "GPT-6 Luna",
     "model_key": "openai/gpt-6-luna",
     "provider": "openai",
     "rank": 20,
     "score": 78.9474
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.4338
     },
     "model": "GPT-5.4",
     "model_key": "openai/gpt-5-4",
     "provider": "openai",
     "rank": 21,
     "score": 78.5965
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.4482
     },
     "model": "Kimi K3.1 Pro",
     "model_key": "moonshotai/kimi-k3-1-pro",
     "provider": "moonshotai",
     "rank": 22,
     "score": 78.2456
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.5664
     },
     "model": "Claude Haiku 5.5",
     "model_key": "anthropic/claude-haiku-5-5",
     "provider": "anthropic",
     "rank": 23,
     "score": 75.0877
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.5784
     },
     "model": "Qwen3.8 Max",
     "model_key": "qwen/qwen3-8-max",
     "provider": "qwen",
     "rank": 24,
     "score": 74.7368
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.5902
     },
     "model": "Muse Spark 1.3",
     "model_key": "meta-llama/muse-spark-1-3",
     "provider": "meta-llama",
     "rank": 25,
     "score": 74.386
    }
   ],
   "source": {
    "attribution": "Epoch AI",
    "licence": "CC-BY-4.0",
    "name": "Epoch AI",
    "url": "https://epoch.ai/benchmarks",
    "via": null
   },
   "unit": "%",
   "updated_at": "2026-11-06"
  },
  {
   "category": "coding",
   "description": "Real GitHub issues the model must fix so the repo's tests pass (Epoch's own runs).",
   "higher_is_better": true,
   "id": "swe-bench-verified",
   "metric": "resolved",
   "name": "SWE-bench Verified",
   "rows": [
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.6901
     },
     "model": "Claude Opus 4.7",
     "model_key": "anthropic/claude-opus-4-7",
     "provider": "anthropic",
     "rank": 1,
     "score": 83.4711
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.8
     },
     "model": "GPT-5.5",
     "model_key": "openai/gpt-5-5",
     "provider": "openai",
     "rank": 2,
     "score": 80.5785
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.8422
     },
     "model": "Gemini 3.5 Flash",
     "model_key": "google/gemini-3-5-flash",
     "provider": "google",
     "rank": 3,
     "score": 79.3388
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.8624
     },
     "model": "Claude Opus 4.6",
     "model_key": "anthropic/claude-opus-4-6",
     "provider": "anthropic",
     "rank": 4,
     "score": 78.719
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.8657
     },
     "model": "GLM-5.2",
     "model_key": "z-ai/glm-5-2",
     "provider": "z-ai",
     "rank": 5,
     "score": 78.7
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.8978
     },
     "model": "DeepSeek v4 Pro",
     "model_key": "deepseek/deepseek-v4-pro",
     "provider": "deepseek",
     "rank": 6,
     "score": 77.6398
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.9068
     },
     "model": "Qwen3.7 Max",
     "model_key": "qwen/qwen3-7-max",
     "provider": "qwen",
     "rank": 7,
     "score": 77.2727
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.9189
     },
     "model": "GPT-5.4",
     "model_key": "openai/gpt-5-4",
     "provider": "openai",
     "rank": 8,
     "score": 76.8595
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.9249
     },
     "model": "Claude Opus 4.5",
     "model_key": "anthropic/claude-opus-4-5",
     "provider": "anthropic",
     "rank": 9,
     "score": 76.6529
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.9249
     },
     "model": "Kimi K2.6",
     "model_key": "moonshotai/kimi-k2-6",
     "provider": "moonshotai",
     "rank": 10,
     "score": 76.6529
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.9249
     },
     "model": "Qwen 3.6 Max",
     "model_key": "qwen/qwen-3-6-max",
     "provider": "qwen",
     "rank": 11,
     "score": 76.6529
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.9537
     },
     "model": "Gemini 3.1 Pro Preview",
     "model_key": "google/gemini-3-1-pro-preview",
     "provider": "google",
     "rank": 12,
     "score": 75.6198
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.9593
     },
     "model": "Gemini 3 Flash Preview",
     "model_key": "google/gemini-3-flash-preview",
     "provider": "google",
     "rank": 13,
     "score": 75.4132
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.9648
     },
     "model": "Claude Sonnet 4.6",
     "model_key": "anthropic/claude-sonnet-4-6",
     "provider": "anthropic",
     "rank": 14,
     "score": 75.2066
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.9757
     },
     "model": "GPT-5.3 Codex",
     "model_key": "openai/gpt-5-3-codex",
     "provider": "openai",
     "rank": 15,
     "score": 74.7934
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.99
     },
     "model": "GLM-5.1",
     "model_key": "z-ai/glm-5-1",
     "provider": "z-ai",
     "rank": 16,
     "score": 74.17
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.0018
     },
     "model": "GPT-5.2",
     "model_key": "openai/gpt-5-2",
     "provider": "openai",
     "rank": 17,
     "score": 73.7603
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.0018
     },
     "model": "Kimi K2.5",
     "model_key": "moonshotai/kimi-k2-5",
     "provider": "moonshotai",
     "rank": 18,
     "score": 73.7603
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.0068
     },
     "model": "GPT-5",
     "model_key": "openai/gpt-5",
     "provider": "openai",
     "rank": 19,
     "score": 73.5537
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.0118
     },
     "model": "claude-opus-4-1-20250805",
     "model_key": "anthropic/claude-opus-4-1",
     "provider": "anthropic",
     "rank": 20,
     "score": 73.3471
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.0216
     },
     "model": "Gemini 3 Pro Preview",
     "model_key": "google/gemini-3-pro-preview",
     "provider": "google",
     "rank": 21,
     "score": 72.9339
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.0894
     },
     "model": "GLM-5",
     "model_key": "z-ai/glm-5",
     "provider": "z-ai",
     "rank": 22,
     "score": 72.0779
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.0587
     },
     "model": "Claude Sonnet 4.5",
     "model_key": "anthropic/claude-sonnet-4-5",
     "provider": "anthropic",
     "rank": 23,
     "score": 71.281
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.0718
     },
     "model": "Claude Opus 4",
     "model_key": "anthropic/claude-opus-4",
     "provider": "anthropic",
     "rank": 24,
     "score": 70.6612
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.123
     },
     "model": "GPT-5.1",
     "model_key": "openai/gpt-5-1",
     "provider": "openai",
     "rank": 25,
     "score": 67.9752
    }
   ],
   "source": {
    "attribution": "Epoch AI",
    "licence": "CC-BY-4.0",
    "name": "Epoch AI",
    "url": "https://epoch.ai/benchmarks/swe-bench-verified",
    "via": null
   },
   "unit": "%",
   "updated_at": "2026-06-25"
  },
  {
   "category": "agentic",
   "description": "Agent completes real tasks in a terminal; best agent harness per model.",
   "higher_is_better": true,
   "id": "terminal-bench",
   "metric": "accuracy",
   "name": "Terminal-Bench",
   "rows": [
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.0892
     },
     "model": "GPT-5.5",
     "model_key": "openai/gpt-5-5",
     "provider": "openai",
     "rank": 1,
     "score": 84.7191
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2
     },
     "model": "GPT-5.4",
     "model_key": "openai/gpt-5-4",
     "provider": "openai",
     "rank": 2,
     "score": 81.8
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.1123
     },
     "model": "Claude Opus 4.7",
     "model_key": "anthropic/claude-opus-4-7",
     "provider": "anthropic",
     "rank": 3,
     "score": 80.2247
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.6243
     },
     "model": "Gemini 3.1 Pro Preview",
     "model_key": "google/gemini-3-1-pro-preview",
     "provider": "google",
     "rank": 4,
     "score": 80.2247
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.6
     },
     "model": "Claude Opus 4.6",
     "model_key": "anthropic/claude-opus-4-6",
     "provider": "anthropic",
     "rank": 5,
     "score": 79.8
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.2022
     },
     "model": "GPT-5.3 Codex",
     "model_key": "openai/gpt-5-3-codex",
     "provider": "openai",
     "rank": 6,
     "score": 78.427
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.0887
     },
     "model": "Gemini 3 Pro Preview",
     "model_key": "google/gemini-3-pro-preview",
     "provider": "google",
     "rank": 7,
     "score": 69.4257
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 3.0621
     },
     "model": "GPT-5.2 Codex",
     "model_key": "openai/gpt-5-2-codex",
     "provider": "openai",
     "rank": 8,
     "score": 66.5169
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.7682
     },
     "model": "GPT-5.2",
     "model_key": "openai/gpt-5-2",
     "provider": "openai",
     "rank": 9,
     "score": 64.9438
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.7856
     },
     "model": "Gemini 3 Flash Preview",
     "model_key": "google/gemini-3-flash-preview",
     "provider": "google",
     "rank": 10,
     "score": 64.2697
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.6972
     },
     "model": "Claude Opus 4.5",
     "model_key": "anthropic/claude-opus-4-5",
     "provider": "anthropic",
     "rank": 11,
     "score": 63.1461
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.8944
     },
     "model": "GPT-5.1 Codex Mini",
     "model_key": "openai/gpt-5-1-codex-mini",
     "provider": "openai",
     "rank": 12,
     "score": 61.573
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.661
     },
     "model": "GPT-5.1-Codex-Max",
     "model_key": "openai/gpt-5-1-codex-max",
     "provider": "openai",
     "rank": 13,
     "score": 60.4494
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.9
     },
     "model": "GPT-5.1 Codex",
     "model_key": "openai/gpt-5-1-codex",
     "provider": "openai",
     "rank": 14,
     "score": 57.8
    },
    {
     "delta_7d": null,
     "extra": {},
     "model": "Grok 4.20 Reasoning",
     "model_key": "x-ai/grok-4-20-reasoning",
     "provider": "x-ai",
     "rank": 15,
     "score": 57.3034
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.83
     },
     "model": "Claude Sonnet 4.6",
     "model_key": "anthropic/claude-sonnet-4-6",
     "provider": "anthropic",
     "rank": 16,
     "score": 53.427
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.6266
     },
     "model": "GLM-5",
     "model_key": "z-ai/glm-5",
     "provider": "z-ai",
     "rank": 17,
     "score": 52.4157
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.8973
     },
     "model": "GPT-5",
     "model_key": "openai/gpt-5",
     "provider": "openai",
     "rank": 18,
     "score": 49.6067
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.8
     },
     "model": "GPT-5.1",
     "model_key": "openai/gpt-5-1",
     "provider": "openai",
     "rank": 19,
     "score": 47.6
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.3514
     },
     "model": "Claude Sonnet 4.5",
     "model_key": "anthropic/claude-sonnet-4-5",
     "provider": "anthropic",
     "rank": 20,
     "score": 46.5169
    },
    {
     "delta_7d": null,
     "extra": {},
     "model": "MiniMax M2.7",
     "model_key": "minimax/minimax-m2-7",
     "provider": "minimax",
     "rank": 21,
     "score": 45.0562
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.7054
     },
     "model": "GPT-5-codex",
     "model_key": "openai/gpt-5-codex",
     "provider": "openai",
     "rank": 22,
     "score": 44.3258
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.8777
     },
     "model": "Kimi K2.5",
     "model_key": "moonshotai/kimi-k2-5",
     "provider": "moonshotai",
     "rank": 23,
     "score": 43.2022
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.9382
     },
     "model": "MiniMax M2.7 Highspeed",
     "model_key": "minimax/minimax-m2-7-highspeed",
     "provider": "minimax",
     "rank": 24,
     "score": 42.9213
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2.8245
     },
     "model": "minimax-m2.5",
     "model_key": "minimax/minimax-m2-5",
     "provider": "minimax",
     "rank": 25,
     "score": 42.6966
    }
   ],
   "source": {
    "attribution": "Terminal-Bench, via Epoch AI",
    "licence": "CC-BY-4.0",
    "name": "Terminal-Bench",
    "url": "https://www.tbench.ai/",
    "via": {
     "name": "Epoch AI",
     "url": "https://epoch.ai/benchmarks/terminal-bench"
    }
   },
   "unit": "%",
   "updated_at": "2026-04-23"
  },
  {
   "category": "reasoning",
   "description": "Very hard expert-written questions across many fields.",
   "higher_is_better": true,
   "id": "hle",
   "metric": "accuracy",
   "name": "Humanity's Last Exam",
   "rows": [
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.94
     },
     "model": "GPT-6 Astra",
     "model_key": "openai/gpt-6-astra",
     "provider": "openai",
     "rank": 1,
     "score": 54.8
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 2
     },
     "model": "Claude Fable 5.1",
     "model_key": "anthropic/claude-fable-5-1",
     "provider": "anthropic",
     "rank": 2,
     "score": 46.5
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.96
     },
     "model": "Gemini 3.1 Pro Preview",
     "model_key": "google/gemini-3-1-pro-preview",
     "provider": "google",
     "rank": 3,
     "score": 46.44
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.95
     },
     "model": "Claude Sonnet 5.5",
     "model_key": "anthropic/claude-sonnet-5-5",
     "provider": "anthropic",
     "rank": 4,
     "score": 46.12
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.96
     },
     "model": "Gemini 3.8 Flash",
     "model_key": "google/gemini-3-8-flash",
     "provider": "google",
     "rank": 5,
     "score": 44.52
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.95
     },
     "model": "GPT-5.4 Pro",
     "model_key": "openai/gpt-5-4-pro",
     "provider": "openai",
     "rank": 6,
     "score": 44.32
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.92
     },
     "model": "Muse Spark",
     "model_key": "meta-llama/muse-spark",
     "provider": "meta-llama",
     "rank": 7,
     "score": 40.56
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.9
     },
     "model": "Gemini 3 Pro Preview",
     "model_key": "google/gemini-3-pro-preview",
     "provider": "google",
     "rank": 8,
     "score": 37.52
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.88
     },
     "model": "GPT-5.4",
     "model_key": "openai/gpt-5-4",
     "provider": "openai",
     "rank": 9,
     "score": 36.24
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.88
     },
     "model": "Claude Opus 4.7",
     "model_key": "anthropic/claude-opus-4-7",
     "provider": "anthropic",
     "rank": 10,
     "score": 36.2
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.86
     },
     "model": "Claude Opus 4.6",
     "model_key": "anthropic/claude-opus-4-6",
     "provider": "anthropic",
     "rank": 11,
     "score": 34.44
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.82
     },
     "model": "GPT-5 Pro",
     "model_key": "openai/gpt-5-pro",
     "provider": "openai",
     "rank": 12,
     "score": 31.64
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.83
     },
     "model": "Claude Haiku 5.5",
     "model_key": "anthropic/claude-haiku-5-5",
     "provider": "anthropic",
     "rank": 13,
     "score": 30.12
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.76
     },
     "model": "GPT-5.2",
     "model_key": "openai/gpt-5-2",
     "provider": "openai",
     "rank": 14,
     "score": 27.8
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.7
     },
     "model": "GPT-5",
     "model_key": "openai/gpt-5",
     "provider": "openai",
     "rank": 15,
     "score": 25.32
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.7
     },
     "model": "Claude Opus 4.5",
     "model_key": "anthropic/claude-opus-4-5",
     "provider": "anthropic",
     "rank": 16,
     "score": 25.2
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.81
     },
     "model": "Kimi K2.5",
     "model_key": "moonshotai/kimi-k2-5",
     "provider": "moonshotai",
     "rank": 17,
     "score": 24.37
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.67
     },
     "model": "GPT-5.1",
     "model_key": "openai/gpt-5-1",
     "provider": "openai",
     "rank": 18,
     "score": 23.68
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.61
     },
     "model": "Gemini 2.5 Pro Preview",
     "model_key": "google/gemini-2-5-pro-preview",
     "provider": "google",
     "rank": 19,
     "score": 21.64
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.58
     },
     "model": "o3",
     "model_key": "openai/o3",
     "provider": "openai",
     "rank": 20,
     "score": 20.32
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.55
     },
     "model": "GPT-5 mini",
     "model_key": "openai/gpt-5-mini",
     "provider": "openai",
     "rank": 21,
     "score": 19.44
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.51
     },
     "model": "Gemini 2.5 Pro Exp",
     "model_key": "google/gemini-2-5-pro-exp",
     "provider": "google",
     "rank": 22,
     "score": 18.16
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.51
     },
     "model": "o4-mini",
     "model_key": "openai/o4-mini",
     "provider": "openai",
     "rank": 23,
     "score": 18.08
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.35
     },
     "model": "Claude Sonnet 4.5",
     "model_key": "anthropic/claude-sonnet-4-5",
     "provider": "anthropic",
     "rank": 24,
     "score": 13.72
    },
    {
     "delta_7d": null,
     "extra": {
      "stderr": 1.28
     },
     "model": "Gemini 2.5 Flash Preview",
     "model_key": "google/gemini-2-5-flash-preview",
     "provider": "google",
     "rank": 25,
     "score": 12.08
    }
   ],
   "source": {
    "attribution": "Humanity's Last Exam (CAIS & Scale AI), via Epoch AI",
    "licence": "CC-BY-4.0",
    "name": "Humanity's Last Exam (CAIS & Scale AI)",
    "url": "https://lastexam.ai/",
    "via": {
     "name": "Epoch AI",
     "url": "https://epoch.ai/benchmarks/hle"
    }
   },
   "unit": "%",
   "updated_at": "2026-10-07"
  },
  {
   "category": "reasoning",
   "description": "Novel visual pattern puzzles that test fluid reasoning.",
   "higher_is_better": true,
   "id": "arc-agi-2",
   "metric": "score",
   "name": "ARC-AGI-2",
   "rows": [
    {
     "delta_7d": null,
     "extra": {},
     "model": "GPT-6 Astra",
     "model_key": "openai/gpt-6-astra",
     "provider": "openai",
     "rank": 1,
     "score": 95
    },
    {
     "delta_7d": null,
     "extra": {},
     "model": "GPT-6.1 Sol",
     "model_key": "openai/gpt-6-1-sol",
     "provider": "openai",
     "rank": 2,
     "score": 94.1667
    },
    {
     "delta_7d": null,
     "extra": {},
     "model": "Claude Fable 5.2",
     "model_key": "anthropic/claude-fable-5-2",
     "provider": "anthropic",
     "rank": 3,
     "score": 93.75
    },
    {
     "delta_7d": null,
     "extra": {},
     "model": "Claude Opus 5.5",
     "model_key": "anthropic/claude-opus-5-5",
     "provider": "anthropic",
     "rank": 4,
     "score": 93.3333
    },
    {
     "delta_7d": null,
     "extra": {
      "cost_per_task_usd": 1.44
     },
     "model": "GPT-5.6 Sol",
     "model_key": "openai/gpt-5-6-sol",
     "provider": "openai",
     "rank": 5,
     "score": 92.5
    },
    {
     "delta_7d": null,
     "extra": {},
     "model": "Gemini 4 Argon",
     "model_key": "google/gemini-4-argon",
     "provider": "google",
     "rank": 6,
     "score": 91.25
    },
    {
     "delta_7d": null,
     "extra": {},
     "model": "Claude Opus 5",
     "model_key": "anthropic/claude-opus-5",
     "provider": "anthropic",
     "rank": 7,
     "score": 90.42
    },
    {
     "delta_7d": null,
     "extra": {},
     "model": "Claude Fable 5.1",
     "model_key": "anthropic/claude-fable-5-1",
     "provider": "anthropic",
     "rank": 8,
     "score": 90
    },
    {
     "delta_7d": null,
     "extra": {},
     "model": "GPT-6 Sol",
     "model_key": "openai/gpt-6-sol",
     "provider": "openai",
     "rank": 9,
     "score": 89.5833
    },
    {
     "delta_7d": null,
     "extra": {},
     "model": "Claude Fable 5",
     "model_key": "anthropic/claude-fable-5",
     "provider": "anthropic",
     "rank": 10,
     "score": 89.1667
    },
    {
     "delta_7d": null,
     "extra": {},
     "model": "Gemini 3.8 Flash",
     "model_key": "google/gemini-3-8-flash",
     "provider": "google",
     "rank": 11,
     "score": 89.1667
    },
    {
     "delta_7d": null,
     "extra": {
      "cost_per_task_usd": 1.87
     },
     "model": "GPT-5.5",
     "model_key": "openai/gpt-5-5",
     "provider": "openai",
     "rank": 12,
     "score": 85
    },
    {
     "delta_7d": null,
     "extra": {},
     "model": "Gemini 3.7 Flash",
     "model_key": "google/gemini-3-7-flash",
     "provider": "google",
     "rank": 13,
     "score": 84.5833
    },
    {
     "delta_7d": null,
     "extra": {},
     "model": "Gemini 3 Deep Think",
     "model_key": "google/gemini-3-deep-think",
     "provider": "google",
     "rank": 14,
     "score": 84.58
    },
    {
     "delta_7d": null,
     "extra": {
      "cost_per_task_usd": 10.51
     },
     "model": "GPT-5.5 Pro",
     "model_key": "openai/gpt-5-5-pro",
     "provider": "openai",
     "rank": 15,
     "score": 84.58
    },
    {
     "delta_7d": null,
     "extra": {
      "cost_per_task_usd": 1.09
     },
     "model": "GPT-5.6 Terra",
     "model_key": "openai/gpt-5-6-terra",
     "provider": "openai",
     "rank": 16,
     "score": 83.9
    },
    {
     "delta_7d": null,
     "extra": {},
     "model": "Claude Sonnet 5.5",
     "model_key": "anthropic/claude-sonnet-5-5",
     "provider": "anthropic",
     "rank": 17,
     "score": 83.8889
    },
    {
     "delta_7d": null,
     "extra": {
      "cost_per_task_usd": 16.41
     },
     "model": "GPT-5.4 Pro",
     "model_key": "openai/gpt-5-4-pro",
     "provider": "openai",
     "rank": 18,
     "score": 83.33
    },
    {
     "delta_7d": null,
     "extra": {},
     "model": "Inkling 2 Small",
     "model_key": "thinking-machines/inkling-2-small",
     "provider": "thinking-machines",
     "rank": 19,
     "score": 81.6667
    },
    {
     "delta_7d": null,
     "extra": {},
     "model": "DeepSeek V4.1 Flash 1017",
     "model_key": "deepseek/deepseek-v4-1-flash-1017",
     "provider": "deepseek",
     "rank": 20,
     "score": 78.0556
    },
    {
     "delta_7d": null,
     "extra": {
      "cost_per_task_usd": 0.9622
     },
     "model": "Gemini 3.1 Pro Preview",
     "model_key": "google/gemini-3-1-pro-preview",
     "provider": "google",
     "rank": 21,
     "score": 77.1
    },
    {
     "delta_7d": null,
     "extra": {},
     "model": "Claude Opus 4.7",
     "model_key": "anthropic/claude-opus-4-7",
     "provider": "anthropic",
     "rank": 22,
     "score": 75.83
    },
    {
     "delta_7d": null,
     "extra": {
      "cost_per_task_usd": 1.52
     },
     "model": "GPT-5.4",
     "model_key": "openai/gpt-5-4",
     "provider": "openai",
     "rank": 23,
     "score": 73.95
    },
    {
     "delta_7d": null,
     "extra": {},
     "model": "DeepSeek V4.1 Flash",
     "model_key": "deepseek/deepseek-v4-1-flash",
     "provider": "deepseek",
     "rank": 24,
     "score": 72.9167
    },
    {
     "delta_7d": null,
     "extra": {
      "cost_per_task_usd": 2.7425
     },
     "model": "Claude Opus 4.8",
     "model_key": "anthropic/claude-opus-4-8",
     "provider": "anthropic",
     "rank": 25,
     "score": 72.08
    }
   ],
   "source": {
    "attribution": "ARC Prize Foundation, via Epoch AI",
    "licence": "CC-BY-4.0",
    "name": "ARC Prize Foundation",
    "url": "https://arcprize.org/",
    "via": {
     "name": "Epoch AI",
     "url": "https://epoch.ai/benchmarks/arc-agi-2"
    }
   },
   "unit": "%",
   "updated_at": "2026-11-09"
  },
  {
   "category": "general",
   "description": "Human preference rating from blind side-by-side chats (style-controlled).",
   "higher_is_better": true,
   "id": "arena-text",
   "metric": "rating",
   "name": "LMArena Text",
   "rows": [
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1535.6,
      "ci_low": 1516.4,
      "source_rank": 1,
      "votes": 4282
     },
     "model": "Gemini 4 Carbon (High)",
     "model_key": "google/gemini-4-carbon",
     "provider": "google",
     "rank": 1,
     "score": 1526
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1527.9,
      "ci_low": 1516.9,
      "source_rank": 2,
      "votes": 19787
     },
     "model": "Gemini 4 Argon (High)",
     "model_key": "google/gemini-4-argon",
     "provider": "google",
     "rank": 2,
     "score": 1522.4
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1517.3,
      "ci_low": 1504.6,
      "source_rank": 3,
      "votes": 12032
     },
     "model": "Muse Wonder (Max)",
     "model_key": "meta-llama/muse-wonder",
     "provider": "meta-llama",
     "rank": 3,
     "score": 1511
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1512.1,
      "ci_low": 1500,
      "source_rank": 4,
      "votes": 15249
     },
     "model": "Claude Opus 5.5 (High)",
     "model_key": "anthropic/claude-opus-5-5",
     "provider": "anthropic",
     "rank": 4,
     "score": 1506.1
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1508.1,
      "ci_low": 1499.7,
      "source_rank": 5,
      "votes": 45524
     },
     "model": "Claude Fable 5 (High)",
     "model_key": "anthropic/claude-fable-5",
     "provider": "anthropic",
     "rank": 5,
     "score": 1503.9
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1506.3,
      "ci_low": 1499.8,
      "source_rank": 6,
      "votes": 84424
     },
     "model": "Claude Opus 4.6 (High)",
     "model_key": "anthropic/claude-opus-4-6",
     "provider": "anthropic",
     "rank": 6,
     "score": 1503
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1508.2,
      "ci_low": 1497.2,
      "source_rank": 7,
      "votes": 20087
     },
     "model": "Claude Fable 5.1 (Max)",
     "model_key": "anthropic/claude-fable-5-1",
     "provider": "anthropic",
     "rank": 7,
     "score": 1502.7
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1504.8,
      "ci_low": 1497.3,
      "source_rank": 8,
      "votes": 64308
     },
     "model": "Claude Opus 4.7 (High)",
     "model_key": "anthropic/claude-opus-4-7",
     "provider": "anthropic",
     "rank": 8,
     "score": 1501.1
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1501.7,
      "ci_low": 1493.3,
      "source_rank": 9,
      "votes": 61143
     },
     "model": "Gemini 3.8 Flash (High)",
     "model_key": "google/gemini-3-8-flash",
     "provider": "google",
     "rank": 9,
     "score": 1497.5
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1501.4,
      "ci_low": 1491.2,
      "source_rank": 11,
      "votes": 26111
     },
     "model": "Muse Spark 1.3 (Max)",
     "model_key": "meta-llama/muse-spark-1-3",
     "provider": "meta-llama",
     "rank": 10,
     "score": 1496.3
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1499.3,
      "ci_low": 1487.6,
      "source_rank": 13,
      "votes": 15004
     },
     "model": "Muse Spark 1.2 (xHigh)",
     "model_key": "meta-llama/muse-spark-1-2",
     "provider": "meta-llama",
     "rank": 11,
     "score": 1493.5
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1494.9,
      "ci_low": 1486.6,
      "source_rank": 14,
      "votes": 45562
     },
     "model": "Muse Spark 1.1",
     "model_key": "meta-llama/muse-spark-1-1",
     "provider": "meta-llama",
     "rank": 12,
     "score": 1490.7
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1493.4,
      "ci_low": 1486.1,
      "source_rank": 15,
      "votes": 73697
     },
     "model": "Claude Opus 5 (High)",
     "model_key": "anthropic/claude-opus-5",
     "provider": "anthropic",
     "rank": 13,
     "score": 1489.8
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1494.1,
      "ci_low": 1484.7,
      "source_rank": 16,
      "votes": 33300
     },
     "model": "Gemini 3.7 Flash (High)",
     "model_key": "google/gemini-3-7-flash",
     "provider": "google",
     "rank": 14,
     "score": 1489.4
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1494.9,
      "ci_low": 1483.3,
      "source_rank": 17,
      "votes": 14009
     },
     "model": "Muse Spark",
     "model_key": "meta-llama/muse-spark",
     "provider": "meta-llama",
     "rank": 15,
     "score": 1489.1
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1491.4,
      "ci_low": 1482.8,
      "source_rank": 19,
      "votes": 40867
     },
     "model": "Kimi K3 (Max)",
     "model_key": "moonshotai/kimi-k3",
     "provider": "moonshotai",
     "rank": 16,
     "score": 1487.1
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1489.7,
      "ci_low": 1483.8,
      "source_rank": 20,
      "votes": 132296
     },
     "model": "Gemini 3.1 Pro Preview",
     "model_key": "google/gemini-3-1-pro-preview",
     "provider": "google",
     "rank": 17,
     "score": 1486.7
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1489.4,
      "ci_low": 1481.8,
      "source_rank": 21,
      "votes": 41829
     },
     "model": "Gemini 3 Pro",
     "model_key": "google/gemini-3-pro",
     "provider": "google",
     "rank": 18,
     "score": 1485.6
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1489.2,
      "ci_low": 1480.7,
      "source_rank": 22,
      "votes": 43994
     },
     "model": "GPT-5.6 Sol (xHigh)",
     "model_key": "openai/gpt-5-6-sol",
     "provider": "openai",
     "rank": 19,
     "score": 1484.9
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1488.6,
      "ci_low": 1479.4,
      "source_rank": 23,
      "votes": 36324
     },
     "model": "Qwen3.8-Max",
     "model_key": "qwen/qwen3-8-max",
     "provider": "qwen",
     "rank": 20,
     "score": 1484
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1493.7,
      "ci_low": 1472.3,
      "source_rank": 24,
      "votes": 3443
     },
     "model": "Kimi K3.1 Pro (Max)",
     "model_key": "moonshotai/kimi-k3-1-pro",
     "provider": "moonshotai",
     "rank": 21,
     "score": 1483
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1487,
      "ci_low": 1478.1,
      "source_rank": 25,
      "votes": 35405
     },
     "model": "Gemini 3.6 Flash (High)",
     "model_key": "google/gemini-3-6-flash",
     "provider": "google",
     "rank": 22,
     "score": 1482.5
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1487.3,
      "ci_low": 1477.8,
      "source_rank": 26,
      "votes": 31017
     },
     "model": "GLM-5.3 (Max)",
     "model_key": "z-ai/glm-5-3",
     "provider": "z-ai",
     "rank": 23,
     "score": 1482.5
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1488.4,
      "ci_low": 1475.7,
      "source_rank": 27,
      "votes": 12952
     },
     "model": "GPT-6.1 Sol (Max)",
     "model_key": "openai/gpt-6-1-sol",
     "provider": "openai",
     "rank": 24,
     "score": 1482.1
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1485.7,
      "ci_low": 1477.9,
      "source_rank": 28,
      "votes": 63128
     },
     "model": "Claude Opus 4.8 (High)",
     "model_key": "anthropic/claude-opus-4-8",
     "provider": "anthropic",
     "rank": 25,
     "score": 1481.8
    }
   ],
   "source": {
    "attribution": "LMArena",
    "licence": "CC-BY-4.0",
    "name": "LMArena",
    "url": "https://arena.ai/leaderboard/text",
    "via": null
   },
   "unit": "rating",
   "updated_at": "2026-11-09"
  },
  {
   "category": "coding",
   "description": "Human preference rating for building web apps from a prompt.",
   "higher_is_better": true,
   "id": "arena-webdev",
   "metric": "rating",
   "name": "LMArena WebDev",
   "rows": [
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1829.8,
      "ci_low": 1805.7,
      "source_rank": 1,
      "votes": 4009
     },
     "model": "Claude Opus 5.5 (Max)",
     "model_key": "anthropic/claude-opus-5-5",
     "provider": "anthropic",
     "rank": 1,
     "score": 1817.7
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1784.6,
      "ci_low": 1764,
      "source_rank": 2,
      "votes": 5493
     },
     "model": "Claude Sonnet 5.5 (xHigh)",
     "model_key": "anthropic/claude-sonnet-5-5",
     "provider": "anthropic",
     "rank": 2,
     "score": 1774.3
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1783.3,
      "ci_low": 1752.7,
      "source_rank": 3,
      "votes": 2128
     },
     "model": "Gemini 4 Carbon (High)",
     "model_key": "google/gemini-4-carbon",
     "provider": "google",
     "rank": 3,
     "score": 1768
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1775.7,
      "ci_low": 1758,
      "source_rank": 4,
      "votes": 8117
     },
     "model": "GPT-6 Astra (Max)",
     "model_key": "openai/gpt-6-astra",
     "provider": "openai",
     "rank": 4,
     "score": 1766.8
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1752.7,
      "ci_low": 1736.4,
      "source_rank": 5,
      "votes": 9453
     },
     "model": "Claude Fable 5.1 (Max)",
     "model_key": "anthropic/claude-fable-5-1",
     "provider": "anthropic",
     "rank": 5,
     "score": 1744.6
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1753,
      "ci_low": 1732.7,
      "source_rank": 6,
      "votes": 5250
     },
     "model": "GPT-6.1 Sol (Max)",
     "model_key": "openai/gpt-6-1-sol",
     "provider": "openai",
     "rank": 6,
     "score": 1742.8
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1738.3,
      "ci_low": 1709.5,
      "source_rank": 7,
      "votes": 2080
     },
     "model": "Muse Wonder Full (Max)",
     "model_key": "meta-llama/muse-wonder-full",
     "provider": "meta-llama",
     "rank": 7,
     "score": 1723.9
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1696.9,
      "ci_low": 1684.7,
      "source_rank": 9,
      "votes": 19593
     },
     "model": "Claude Opus 5 (Max)",
     "model_key": "anthropic/claude-opus-5",
     "provider": "anthropic",
     "rank": 8,
     "score": 1690.8
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1697.9,
      "ci_low": 1677.8,
      "source_rank": 10,
      "votes": 4591
     },
     "model": "Kimi K3.1 Pro (Max)",
     "model_key": "moonshotai/kimi-k3-1-pro",
     "provider": "moonshotai",
     "rank": 9,
     "score": 1687.8
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1692.6,
      "ci_low": 1672.6,
      "source_rank": 11,
      "votes": 4626
     },
     "model": "Gemini 4 Argon (High)",
     "model_key": "google/gemini-4-argon",
     "provider": "google",
     "rank": 10,
     "score": 1682.6
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1690,
      "ci_low": 1667.5,
      "source_rank": 12,
      "votes": 3559
     },
     "model": "GPT-6.1 Terra (Max)",
     "model_key": "openai/gpt-6-1-terra",
     "provider": "openai",
     "rank": 11,
     "score": 1678.7
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1683,
      "ci_low": 1670.4,
      "source_rank": 13,
      "votes": 21354
     },
     "model": "Qwen3.8-Max (0902)",
     "model_key": "qwen/qwen3-8-max",
     "provider": "qwen",
     "rank": 12,
     "score": 1676.7
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1682.4,
      "ci_low": 1665.3,
      "source_rank": 14,
      "votes": 6882
     },
     "model": "GPT-6 Sol (Max)",
     "model_key": "openai/gpt-6-sol",
     "provider": "openai",
     "rank": 13,
     "score": 1673.9
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1671,
      "ci_low": 1657.4,
      "source_rank": 16,
      "votes": 13241
     },
     "model": "DeepSeek-V4.1-Flash-20261017 (Max)",
     "model_key": "deepseek/deepseek-v4-1-flash",
     "provider": "deepseek",
     "rank": 14,
     "score": 1664.2
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1662.8,
      "ci_low": 1651.3,
      "source_rank": 17,
      "votes": 22114
     },
     "model": "Kimi K3 (Max)",
     "model_key": "moonshotai/kimi-k3",
     "provider": "moonshotai",
     "rank": 15,
     "score": 1657.1
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1661.4,
      "ci_low": 1647.2,
      "source_rank": 19,
      "votes": 11758
     },
     "model": "Muse Spark 1.3 (Max)",
     "model_key": "meta-llama/muse-spark-1-3",
     "provider": "meta-llama",
     "rank": 16,
     "score": 1654.3
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1660,
      "ci_low": 1640.5,
      "source_rank": 20,
      "votes": 4875
     },
     "model": "Muse Wonder (Max)",
     "model_key": "meta-llama/muse-wonder",
     "provider": "meta-llama",
     "rank": 17,
     "score": 1650.3
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1669.2,
      "ci_low": 1629.8,
      "source_rank": 21,
      "votes": 964
     },
     "model": "Inkling 2 Small",
     "model_key": "thinky/inkling-2-small",
     "provider": "thinky",
     "rank": 18,
     "score": 1649.5
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1647.4,
      "ci_low": 1631.3,
      "source_rank": 22,
      "votes": 7816
     },
     "model": "Grok 4.7 (xHigh)",
     "model_key": "x-ai/grok-4-7",
     "provider": "x-ai",
     "rank": 19,
     "score": 1639.4
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1644.4,
      "ci_low": 1628.7,
      "source_rank": 23,
      "votes": 8576
     },
     "model": "MiMo-V2.6-Flash",
     "model_key": "xiaomi/mimo-v2-6-flash",
     "provider": "xiaomi",
     "rank": 20,
     "score": 1636.5
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1641.3,
      "ci_low": 1621.7,
      "source_rank": 24,
      "votes": 4606
     },
     "model": "Hy4 preview",
     "model_key": "tencent/hy4-preview",
     "provider": "tencent",
     "rank": 21,
     "score": 1631.5
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1631.2,
      "ci_low": 1619.2,
      "source_rank": 25,
      "votes": 17017
     },
     "model": "Claude Fable 5 (High)",
     "model_key": "anthropic/claude-fable-5",
     "provider": "anthropic",
     "rank": 22,
     "score": 1625.2
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1631.7,
      "ci_low": 1617.2,
      "source_rank": 26,
      "votes": 10939
     },
     "model": "Qwen3.8-Flash-Next",
     "model_key": "qwen/qwen3-8-flash-next",
     "provider": "qwen",
     "rank": 23,
     "score": 1624.5
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1631.7,
      "ci_low": 1616.1,
      "source_rank": 27,
      "votes": 8402
     },
     "model": "MiMo-V2.6-Pro",
     "model_key": "xiaomi/mimo-v2-6-pro",
     "provider": "xiaomi",
     "rank": 24,
     "score": 1623.9
    },
    {
     "delta_7d": null,
     "extra": {
      "ci_high": 1629.1,
      "ci_low": 1615.2,
      "source_rank": 28,
      "votes": 12263
     },
     "model": "GLM-5.3 (Max)",
     "model_key": "z-ai/glm-5-3",
     "provider": "z-ai",
     "rank": 25,
     "score": 1622.2
    }
   ],
   "source": {
    "attribution": "LMArena",
    "licence": "CC-BY-4.0",
    "name": "LMArena",
    "url": "https://arena.ai/leaderboard/webdev",
    "via": null
   },
   "unit": "rating",
   "updated_at": "2026-11-09"
  }
 ],
 "generated_at": "2026-11-10T17:18:37Z",
 "models": [
  {
   "context": null,
   "model": "Claude Fable 5.1",
   "model_key": "anthropic/claude-fable-5-1",
   "price_in": null,
   "price_out": null,
   "provider": "anthropic",
   "scores": {
    "arc-agi-2": 90,
    "arena-text": 1502.7,
    "arena-webdev": 1744.6,
    "epoch-eci": 167.04,
    "frontiermath": 90.1754,
    "hle": 46.5
   }
  },
  {
   "context": null,
   "model": "Claude Opus 5",
   "model_key": "anthropic/claude-opus-5",
   "price_in": null,
   "price_out": null,
   "provider": "anthropic",
   "scores": {
    "arc-agi-2": 90.42,
    "arena-text": 1489.8,
    "arena-webdev": 1690.8,
    "epoch-eci": 164.48,
    "frontiermath": 85.614,
    "gpqa-diamond": 93.8763
   }
  },
  {
   "context": null,
   "model": "Claude Sonnet 5.5",
   "model_key": "anthropic/claude-sonnet-5-5",
   "price_in": null,
   "price_out": null,
   "provider": "anthropic",
   "scores": {
    "arc-agi-2": 83.8889,
    "arena-webdev": 1774.3,
    "epoch-eci": 166.32,
    "frontiermath": 88.7719,
    "gpqa-diamond": 95.5808,
    "hle": 46.12
   }
  },
  {
   "context": null,
   "model": "Gemini 3.1 Pro Preview",
   "model_key": "google/gemini-3-1-pro-preview",
   "price_in": null,
   "price_out": null,
   "provider": "google",
   "scores": {
    "arc-agi-2": 77.1,
    "arena-text": 1486.7,
    "gpqa-diamond": 94.4444,
    "hle": 46.44,
    "swe-bench-verified": 75.6198,
    "terminal-bench": 80.2247
   }
  },
  {
   "context": null,
   "model": "Gemini 4 Argon",
   "model_key": "google/gemini-4-argon",
   "price_in": null,
   "price_out": null,
   "provider": "google",
   "scores": {
    "arc-agi-2": 91.25,
    "arena-text": 1522.4,
    "arena-webdev": 1682.6,
    "epoch-eci": 166.53,
    "frontiermath": 85.9649,
    "gpqa-diamond": 96.4015
   }
  },
  {
   "context": null,
   "model": "GPT-5.4",
   "model_key": "openai/gpt-5-4",
   "price_in": null,
   "price_out": null,
   "provider": "openai",
   "scores": {
    "arc-agi-2": 73.95,
    "frontiermath": 78.5965,
    "gpqa-diamond": 93.3,
    "hle": 36.24,
    "swe-bench-verified": 76.8595,
    "terminal-bench": 81.8
   }
  },
  {
   "context": null,
   "model": "GPT-5.5",
   "model_key": "openai/gpt-5-5",
   "price_in": null,
   "price_out": null,
   "provider": "openai",
   "scores": {
    "arc-agi-2": 85,
    "epoch-eci": 160.06,
    "frontiermath": 85.2632,
    "gpqa-diamond": 94.0025,
    "swe-bench-verified": 80.5785,
    "terminal-bench": 84.7191
   }
  },
  {
   "context": null,
   "model": "GPT-6.1 Sol",
   "model_key": "openai/gpt-6-1-sol",
   "price_in": null,
   "price_out": null,
   "provider": "openai",
   "scores": {
    "arc-agi-2": 94.1667,
    "arena-text": 1482.1,
    "arena-webdev": 1742.8,
    "epoch-eci": 168.87,
    "frontiermath": 93.6842,
    "gpqa-diamond": 95.3914
   }
  },
  {
   "context": null,
   "model": "GPT-6 Astra",
   "model_key": "openai/gpt-6-astra",
   "price_in": null,
   "price_out": null,
   "provider": "openai",
   "scores": {
    "arc-agi-2": 95,
    "arena-webdev": 1766.8,
    "epoch-eci": 169.16,
    "frontiermath": 93.6842,
    "gpqa-diamond": 95.7702,
    "hle": 54.8
   }
  },
  {
   "context": null,
   "model": "Claude Fable 5",
   "model_key": "anthropic/claude-fable-5",
   "price_in": null,
   "price_out": null,
   "provider": "anthropic",
   "scores": {
    "arc-agi-2": 89.1667,
    "arena-text": 1503.9,
    "arena-webdev": 1625.2,
    "epoch-eci": 164.25,
    "frontiermath": 87.0175
   }
  },
  {
   "context": null,
   "model": "Claude Opus 4.7",
   "model_key": "anthropic/claude-opus-4-7",
   "price_in": null,
   "price_out": null,
   "provider": "anthropic",
   "scores": {
    "arc-agi-2": 75.83,
    "arena-text": 1501.1,
    "hle": 36.2,
    "swe-bench-verified": 83.4711,
    "terminal-bench": 80.2247
   }
  },
  {
   "context": null,
   "model": "Claude Opus 5.5",
   "model_key": "anthropic/claude-opus-5-5",
   "price_in": null,
   "price_out": null,
   "provider": "anthropic",
   "scores": {
    "arc-agi-2": 93.3333,
    "arena-text": 1506.1,
    "arena-webdev": 1817.7,
    "epoch-eci": 168.67,
    "frontiermath": 91.2281
   }
  },
  {
   "context": null,
   "model": "Gemini 3.8 Flash",
   "model_key": "google/gemini-3-8-flash",
   "price_in": null,
   "price_out": null,
   "provider": "google",
   "scores": {
    "arc-agi-2": 89.1667,
    "arena-text": 1497.5,
    "epoch-eci": 158.66,
    "gpqa-diamond": 95.3914,
    "hle": 44.52
   }
  },
  {
   "context": null,
   "model": "Kimi-K3.1-Pro",
   "model_key": "moonshotai/kimi-k3-1-pro",
   "price_in": null,
   "price_out": null,
   "provider": "moonshotai",
   "scores": {
    "arena-text": 1483,
    "arena-webdev": 1687.8,
    "epoch-eci": 162.02,
    "frontiermath": 78.2456,
    "gpqa-diamond": 94.1919
   }
  },
  {
   "context": null,
   "model": "GPT-5.4 Pro",
   "model_key": "openai/gpt-5-4-pro",
   "price_in": null,
   "price_out": null,
   "provider": "openai",
   "scores": {
    "arc-agi-2": 83.33,
    "epoch-eci": 160.17,
    "frontiermath": 82.4561,
    "gpqa-diamond": 94.6,
    "hle": 44.32
   }
  },
  {
   "context": null,
   "model": "GPT-5.6 Sol",
   "model_key": "openai/gpt-5-6-sol",
   "price_in": null,
   "price_out": null,
   "provider": "openai",
   "scores": {
    "arc-agi-2": 92.5,
    "arena-text": 1484.9,
    "epoch-eci": 163.02,
    "frontiermath": 89.1228,
    "gpqa-diamond": 93.4975
   }
  },
  {
   "context": null,
   "model": "GPT-6 Sol",
   "model_key": "openai/gpt-6-sol",
   "price_in": null,
   "price_out": null,
   "provider": "openai",
   "scores": {
    "arc-agi-2": 89.5833,
    "arena-webdev": 1673.9,
    "epoch-eci": 164,
    "frontiermath": 89.8246,
    "gpqa-diamond": 94.3182
   }
  },
  {
   "context": 1000000,
   "model": "Claude Fable 5.2",
   "model_key": "anthropic/claude-fable-5-2",
   "price_in": 8,
   "price_out": 40,
   "provider": "anthropic",
   "scores": {
    "arc-agi-2": 93.75,
    "epoch-eci": 174.24,
    "frontiermath": 95.0877,
    "gpqa-diamond": 93.8763
   }
  },
  {
   "context": null,
   "model": "Claude Opus 4.6",
   "model_key": "anthropic/claude-opus-4-6",
   "price_in": null,
   "price_out": null,
   "provider": "anthropic",
   "scores": {
    "arena-text": 1503,
    "hle": 34.44,
    "swe-bench-verified": 78.719,
    "terminal-bench": 79.8
   }
  },
  {
   "context": null,
   "model": "Claude Opus 4.8",
   "model_key": "anthropic/claude-opus-4-8",
   "price_in": null,
   "price_out": null,
   "provider": "anthropic",
   "scores": {
    "arc-agi-2": 72.08,
    "arena-text": 1481.8,
    "epoch-eci": 159.91,
    "frontiermath": 80
   }
  },
  {
   "context": null,
   "model": "Gemini 3.7 Flash",
   "model_key": "google/gemini-3-7-flash",
   "price_in": null,
   "price_out": null,
   "provider": "google",
   "scores": {
    "arc-agi-2": 84.5833,
    "arena-text": 1489.4,
    "epoch-eci": 158.44,
    "gpqa-diamond": 94.8232
   }
  },
  {
   "context": null,
   "model": "Muse Spark 1.3",
   "model_key": "meta-llama/muse-spark-1-3",
   "price_in": null,
   "price_out": null,
   "provider": "meta-llama",
   "scores": {
    "arena-text": 1496.3,
    "arena-webdev": 1654.3,
    "epoch-eci": 157.97,
    "frontiermath": 74.386
   }
  },
  {
   "context": null,
   "model": "Muse Wonder",
   "model_key": "meta-llama/muse-wonder",
   "price_in": null,
   "price_out": null,
   "provider": "meta-llama",
   "scores": {
    "arena-text": 1511,
    "arena-webdev": 1650.3,
    "epoch-eci": 162.01,
    "frontiermath": 82.1053
   }
  },
  {
   "context": null,
   "model": "Kimi K3",
   "model_key": "moonshotai/kimi-k3",
   "price_in": null,
   "price_out": null,
   "provider": "moonshotai",
   "scores": {
    "arena-text": 1487.1,
    "arena-webdev": 1657.1,
    "epoch-eci": 158.53,
    "gpqa-diamond": 93.1187
   }
  },
  {
   "context": null,
   "model": "GPT-5.5 Pro",
   "model_key": "openai/gpt-5-5-pro",
   "price_in": null,
   "price_out": null,
   "provider": "openai",
   "scores": {
    "arc-agi-2": 84.58,
    "epoch-eci": 162.75,
    "frontiermath": 87.7193,
    "gpqa-diamond": 93.9214
   }
  },
  {
   "context": null,
   "model": "GPT-5.6 Terra",
   "model_key": "openai/gpt-5-6-terra",
   "price_in": null,
   "price_out": null,
   "provider": "openai",
   "scores": {
    "arc-agi-2": 83.9,
    "epoch-eci": 160.86,
    "frontiermath": 85.9649,
    "gpqa-diamond": 93.3081
   }
  },
  {
   "context": null,
   "model": "GPT-6.1 Terra",
   "model_key": "openai/gpt-6-1-terra",
   "price_in": null,
   "price_out": null,
   "provider": "openai",
   "scores": {
    "arena-webdev": 1678.7,
    "epoch-eci": 168.11,
    "frontiermath": 90.8772,
    "gpqa-diamond": 94.3182
   }
  },
  {
   "context": null,
   "model": "Claude Haiku 5.5",
   "model_key": "anthropic/claude-haiku-5-5",
   "price_in": null,
   "price_out": null,
   "provider": "anthropic",
   "scores": {
    "epoch-eci": 160.25,
    "frontiermath": 75.0877,
    "hle": 30.12
   }
  },
  {
   "context": null,
   "model": "Claude Opus 4.5",
   "model_key": "anthropic/claude-opus-4-5",
   "price_in": null,
   "price_out": null,
   "provider": "anthropic",
   "scores": {
    "hle": 25.2,
    "swe-bench-verified": 76.6529,
    "terminal-bench": 63.1461
   }
  },
  {
   "context": null,
   "model": "Claude Sonnet 4.5",
   "model_key": "anthropic/claude-sonnet-4-5",
   "price_in": null,
   "price_out": null,
   "provider": "anthropic",
   "scores": {
    "hle": 13.72,
    "swe-bench-verified": 71.281,
    "terminal-bench": 46.5169
   }
  },
  {
   "context": null,
   "model": "DeepSeek V4.1 Flash 1017",
   "model_key": "deepseek/deepseek-v4-1-flash-1017",
   "price_in": null,
   "price_out": null,
   "provider": "deepseek",
   "scores": {
    "arc-agi-2": 78.0556,
    "epoch-eci": 159.34,
    "gpqa-diamond": 93.1818
   }
  },
  {
   "context": null,
   "model": "Gemini 3 Pro Preview",
   "model_key": "google/gemini-3-pro-preview",
   "price_in": null,
   "price_out": null,
   "provider": "google",
   "scores": {
    "hle": 37.52,
    "swe-bench-verified": 72.9339,
    "terminal-bench": 69.4257
   }
  },
  {
   "context": null,
   "model": "Kimi K2.5",
   "model_key": "moonshotai/kimi-k2-5",
   "price_in": null,
   "price_out": null,
   "provider": "moonshotai",
   "scores": {
    "hle": 24.37,
    "swe-bench-verified": 73.7603,
    "terminal-bench": 43.2022
   }
  },
  {
   "context": null,
   "model": "GPT-5",
   "model_key": "openai/gpt-5",
   "price_in": null,
   "price_out": null,
   "provider": "openai",
   "scores": {
    "hle": 25.32,
    "swe-bench-verified": 73.5537,
    "terminal-bench": 49.6067
   }
  },
  {
   "context": null,
   "model": "GPT-5.1",
   "model_key": "openai/gpt-5-1",
   "price_in": null,
   "price_out": null,
   "provider": "openai",
   "scores": {
    "hle": 23.68,
    "swe-bench-verified": 67.9752,
    "terminal-bench": 47.6
   }
  },
  {
   "context": null,
   "model": "GPT-5.2",
   "model_key": "openai/gpt-5-2",
   "price_in": null,
   "price_out": null,
   "provider": "openai",
   "scores": {
    "hle": 27.8,
    "swe-bench-verified": 73.7603,
    "terminal-bench": 64.9438
   }
  },
  {
   "context": null,
   "model": "Qwen3.8 Max",
   "model_key": "qwen/qwen3-8-max",
   "price_in": null,
   "price_out": null,
   "provider": "qwen",
   "scores": {
    "arena-text": 1484,
    "arena-webdev": 1676.7,
    "frontiermath": 74.7368
   }
  },
  {
   "context": null,
   "model": "Claude Sonnet 4.6",
   "model_key": "anthropic/claude-sonnet-4-6",
   "price_in": null,
   "price_out": null,
   "provider": "anthropic",
   "scores": {
    "swe-bench-verified": 75.2066,
    "terminal-bench": 53.427
   }
  },
  {
   "context": null,
   "model": "DeepSeek V4.1 Flash",
   "model_key": "deepseek/deepseek-v4-1-flash",
   "price_in": null,
   "price_out": null,
   "provider": "deepseek",
   "scores": {
    "arc-agi-2": 72.9167,
    "arena-webdev": 1664.2
   }
  },
  {
   "context": null,
   "model": "Gemini 3.5 Flash",
   "model_key": "google/gemini-3-5-flash",
   "price_in": null,
   "price_out": null,
   "provider": "google",
   "scores": {
    "gpqa-diamond": 92.803,
    "swe-bench-verified": 79.3388
   }
  },
  {
   "context": null,
   "model": "Gemini 3.6 Flash",
   "model_key": "google/gemini-3-6-flash",
   "price_in": null,
   "price_out": null,
   "provider": "google",
   "scores": {
    "arena-text": 1482.5,
    "gpqa-diamond": 94.1288
   }
  },
  {
   "context": null,
   "model": "Gemini 3 Flash Preview",
   "model_key": "google/gemini-3-flash-preview",
   "price_in": null,
   "price_out": null,
   "provider": "google",
   "scores": {
    "swe-bench-verified": 75.4132,
    "terminal-bench": 64.2697
   }
  },
  {
   "context": 1048576,
   "model": "Gemini 4 Carbon",
   "model_key": "google/gemini-4-carbon",
   "price_in": 2,
   "price_out": 10,
   "provider": "google",
   "scores": {
    "arena-text": 1526,
    "arena-webdev": 1768
   }
  },
  {
   "context": null,
   "model": "Muse Spark",
   "model_key": "meta-llama/muse-spark",
   "price_in": null,
   "price_out": null,
   "provider": "meta-llama",
   "scores": {
    "arena-text": 1489.1,
    "hle": 40.56
   }
  },
  {
   "context": null,
   "model": "GPT-5.3 Codex",
   "model_key": "openai/gpt-5-3-codex",
   "price_in": null,
   "price_out": null,
   "provider": "openai",
   "scores": {
    "swe-bench-verified": 74.7934,
    "terminal-bench": 78.427
   }
  },
  {
   "context": null,
   "model": "GLM-5",
   "model_key": "z-ai/glm-5",
   "price_in": null,
   "price_out": null,
   "provider": "z-ai",
   "scores": {
    "swe-bench-verified": 72.0779,
    "terminal-bench": 52.4157
   }
  },
  {
   "context": null,
   "model": "GLM-5.3 (Max)",
   "model_key": "z-ai/glm-5-3",
   "price_in": null,
   "price_out": null,
   "provider": "z-ai",
   "scores": {
    "arena-text": 1482.5,
    "arena-webdev": 1622.2
   }
  },
  {
   "context": null,
   "model": "Claude Opus 4",
   "model_key": "anthropic/claude-opus-4",
   "price_in": null,
   "price_out": null,
   "provider": "anthropic",
   "scores": {
    "swe-bench-verified": 70.6612
   }
  },
  {
   "context": null,
   "model": "claude-opus-4-1-20250805",
   "model_key": "anthropic/claude-opus-4-1",
   "price_in": null,
   "price_out": null,
   "provider": "anthropic",
   "scores": {
    "swe-bench-verified": 73.3471
   }
  },
  {
   "context": null,
   "model": "Doubao Seed Evolving",
   "model_key": "bytedance/doubao-seed-evolving",
   "price_in": null,
   "price_out": null,
   "provider": "bytedance",
   "scores": {
    "gpqa-diamond": 93.1187
   }
  },
  {
   "context": null,
   "model": "DeepSeek v4 Pro",
   "model_key": "deepseek/deepseek-v4-pro",
   "price_in": null,
   "price_out": null,
   "provider": "deepseek",
   "scores": {
    "swe-bench-verified": 77.6398
   }
  },
  {
   "context": null,
   "model": "Gemini 2.5 Flash Preview",
   "model_key": "google/gemini-2-5-flash-preview",
   "price_in": null,
   "price_out": null,
   "provider": "google",
   "scores": {
    "hle": 12.08
   }
  },
  {
   "context": null,
   "model": "Gemini 2.5 Pro Exp",
   "model_key": "google/gemini-2-5-pro-exp",
   "price_in": null,
   "price_out": null,
   "provider": "google",
   "scores": {
    "hle": 18.16
   }
  },
  {
   "context": null,
   "model": "Gemini 2.5 Pro Preview",
   "model_key": "google/gemini-2-5-pro-preview",
   "price_in": null,
   "price_out": null,
   "provider": "google",
   "scores": {
    "hle": 21.64
   }
  },
  {
   "context": null,
   "model": "Gemini 3 Deep Think",
   "model_key": "google/gemini-3-deep-think",
   "price_in": null,
   "price_out": null,
   "provider": "google",
   "scores": {
    "arc-agi-2": 84.58
   }
  },
  {
   "context": null,
   "model": "Gemini 3 Pro",
   "model_key": "google/gemini-3-pro",
   "price_in": null,
   "price_out": null,
   "provider": "google",
   "scores": {
    "arena-text": 1485.6
   }
  },
  {
   "context": null,
   "model": "Muse Spark 1.1",
   "model_key": "meta-llama/muse-spark-1-1",
   "price_in": null,
   "price_out": null,
   "provider": "meta-llama",
   "scores": {
    "arena-text": 1490.7
   }
  },
  {
   "context": null,
   "model": "Muse Spark 1.2 (xHigh)",
   "model_key": "meta-llama/muse-spark-1-2",
   "price_in": null,
   "price_out": null,
   "provider": "meta-llama",
   "scores": {
    "arena-text": 1493.5
   }
  },
  {
   "context": null,
   "model": "Muse Wonder Full (Max)",
   "model_key": "meta-llama/muse-wonder-full",
   "price_in": null,
   "price_out": null,
   "provider": "meta-llama",
   "scores": {
    "arena-webdev": 1723.9
   }
  },
  {
   "context": null,
   "model": "minimax-m2.5",
   "model_key": "minimax/minimax-m2-5",
   "price_in": null,
   "price_out": null,
   "provider": "minimax",
   "scores": {
    "terminal-bench": 42.6966
   }
  },
  {
   "context": null,
   "model": "MiniMax M2.7",
   "model_key": "minimax/minimax-m2-7",
   "price_in": null,
   "price_out": null,
   "provider": "minimax",
   "scores": {
    "terminal-bench": 45.0562
   }
  },
  {
   "context": null,
   "model": "MiniMax M2.7 Highspeed",
   "model_key": "minimax/minimax-m2-7-highspeed",
   "price_in": null,
   "price_out": null,
   "provider": "minimax",
   "scores": {
    "terminal-bench": 42.9213
   }
  },
  {
   "context": null,
   "model": "Kimi K2.6",
   "model_key": "moonshotai/kimi-k2-6",
   "price_in": null,
   "price_out": null,
   "provider": "moonshotai",
   "scores": {
    "swe-bench-verified": 76.6529
   }
  },
  {
   "context": null,
   "model": "GPT-5.1 Codex",
   "model_key": "openai/gpt-5-1-codex",
   "price_in": null,
   "price_out": null,
   "provider": "openai",
   "scores": {
    "terminal-bench": 57.8
   }
  },
  {
   "context": null,
   "model": "GPT-5.1-Codex-Max",
   "model_key": "openai/gpt-5-1-codex-max",
   "price_in": null,
   "price_out": null,
   "provider": "openai",
   "scores": {
    "terminal-bench": 60.4494
   }
  },
  {
   "context": null,
   "model": "GPT-5.1 Codex Mini",
   "model_key": "openai/gpt-5-1-codex-mini",
   "price_in": null,
   "price_out": null,
   "provider": "openai",
   "scores": {
    "terminal-bench": 61.573
   }
  },
  {
   "context": null,
   "model": "GPT-5.2 Codex",
   "model_key": "openai/gpt-5-2-codex",
   "price_in": null,
   "price_out": null,
   "provider": "openai",
   "scores": {
    "terminal-bench": 66.5169
   }
  },
  {
   "context": null,
   "model": "GPT-5.6 Luna",
   "model_key": "openai/gpt-5-6-luna",
   "price_in": null,
   "price_out": null,
   "provider": "openai",
   "scores": {
    "frontiermath": 82.1053
   }
  },
  {
   "context": null,
   "model": "GPT-5-codex",
   "model_key": "openai/gpt-5-codex",
   "price_in": null,
   "price_out": null,
   "provider": "openai",
   "scores": {
    "terminal-bench": 44.3258
   }
  },
  {
   "context": null,
   "model": "GPT-5 mini",
   "model_key": "openai/gpt-5-mini",
   "price_in": null,
   "price_out": null,
   "provider": "openai",
   "scores": {
    "hle": 19.44
   }
  },
  {
   "context": null,
   "model": "GPT-5 Pro",
   "model_key": "openai/gpt-5-pro",
   "price_in": null,
   "price_out": null,
   "provider": "openai",
   "scores": {
    "hle": 31.64
   }
  },
  {
   "context": null,
   "model": "GPT-6 Luna",
   "model_key": "openai/gpt-6-luna",
   "price_in": null,
   "price_out": null,
   "provider": "openai",
   "scores": {
    "frontiermath": 78.9474
   }
  },
  {
   "context": null,
   "model": "o3",
   "model_key": "openai/o3",
   "price_in": null,
   "price_out": null,
   "provider": "openai",
   "scores": {
    "hle": 20.32
   }
  },
  {
   "context": null,
   "model": "o4-mini",
   "model_key": "openai/o4-mini",
   "price_in": null,
   "price_out": null,
   "provider": "openai",
   "scores": {
    "hle": 18.08
   }
  },
  {
   "context": null,
   "model": "Qwen 3.6 Max",
   "model_key": "qwen/qwen-3-6-max",
   "price_in": null,
   "price_out": null,
   "provider": "qwen",
   "scores": {
    "swe-bench-verified": 76.6529
   }
  },
  {
   "context": null,
   "model": "Qwen3.7 Max",
   "model_key": "qwen/qwen3-7-max",
   "price_in": null,
   "price_out": null,
   "provider": "qwen",
   "scores": {
    "swe-bench-verified": 77.2727
   }
  },
  {
   "context": null,
   "model": "Qwen3.8-Flash-Next",
   "model_key": "qwen/qwen3-8-flash-next",
   "price_in": null,
   "price_out": null,
   "provider": "qwen",
   "scores": {
    "arena-webdev": 1624.5
   }
  },
  {
   "context": null,
   "model": "Hy4 preview",
   "model_key": "tencent/hy4-preview",
   "price_in": null,
   "price_out": null,
   "provider": "tencent",
   "scores": {
    "arena-webdev": 1631.5
   }
  },
  {
   "context": null,
   "model": "Inkling 2 Small",
   "model_key": "thinking-machines/inkling-2-small",
   "price_in": null,
   "price_out": null,
   "provider": "thinking-machines",
   "scores": {
    "arc-agi-2": 81.6667
   }
  },
  {
   "context": null,
   "model": "Inkling 2 Small",
   "model_key": "thinky/inkling-2-small",
   "price_in": null,
   "price_out": null,
   "provider": "thinky",
   "scores": {
    "arena-webdev": 1649.5
   }
  },
  {
   "context": null,
   "model": "Grok 4.20 Reasoning",
   "model_key": "x-ai/grok-4-20-reasoning",
   "price_in": null,
   "price_out": null,
   "provider": "x-ai",
   "scores": {
    "terminal-bench": 57.3034
   }
  },
  {
   "context": null,
   "model": "Grok 4.5",
   "model_key": "x-ai/grok-4-5",
   "price_in": null,
   "price_out": null,
   "provider": "x-ai",
   "scores": {
    "gpqa-diamond": 93.4343
   }
  },
  {
   "context": null,
   "model": "Grok 4.6",
   "model_key": "x-ai/grok-4-6",
   "price_in": null,
   "price_out": null,
   "provider": "x-ai",
   "scores": {
    "gpqa-diamond": 94.0025
   }
  },
  {
   "context": null,
   "model": "Grok 4.7 (xHigh)",
   "model_key": "x-ai/grok-4-7",
   "price_in": null,
   "price_out": null,
   "provider": "x-ai",
   "scores": {
    "arena-webdev": 1639.4
   }
  },
  {
   "context": null,
   "model": "MiMo-V2.6-Flash",
   "model_key": "xiaomi/mimo-v2-6-flash",
   "price_in": null,
   "price_out": null,
   "provider": "xiaomi",
   "scores": {
    "arena-webdev": 1636.5
   }
  },
  {
   "context": null,
   "model": "MiMo-V2.6-Pro",
   "model_key": "xiaomi/mimo-v2-6-pro",
   "price_in": null,
   "price_out": null,
   "provider": "xiaomi",
   "scores": {
    "arena-webdev": 1623.9
   }
  },
  {
   "context": null,
   "model": "GLM-5.1",
   "model_key": "z-ai/glm-5-1",
   "price_in": null,
   "price_out": null,
   "provider": "z-ai",
   "scores": {
    "swe-bench-verified": 74.17
   }
  },
  {
   "context": null,
   "model": "GLM-5.2",
   "model_key": "z-ai/glm-5-2",
   "price_in": null,
   "price_out": null,
   "provider": "z-ai",
   "scores": {
    "swe-bench-verified": 78.7
   }
  }
 ],
 "schema": 1
}