{
  "count": 6,
  "data": [
    {
      "benchmark": "GPQA Diamond (Epoch AI run)",
      "count": 77,
      "scores": [
        {
          "modelId": "gpt-6-astra",
          "name": "GPT-6 Astra",
          "providerId": "openai",
          "score": 95.8,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gemini-3-8-flash",
          "name": "Gemini 3.8 Flash",
          "providerId": "google",
          "score": 95.4,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gemini-3-7-flash",
          "name": "Gemini 3.7 Flash",
          "providerId": "google",
          "score": 94.8,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gpt-5-4-pro",
          "name": "GPT-5.4 Pro",
          "providerId": "openai",
          "score": 94.6,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gemini-3-1-pro-preview",
          "name": "Gemini 3.1 Pro Preview",
          "providerId": "google",
          "score": 94.4,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gemini-3-6-flash",
          "name": "Gemini 3.6 Flash",
          "providerId": "google",
          "score": 94.1,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "grok-4-6",
          "name": "Grok 4.6",
          "providerId": "xai",
          "score": 94,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "claude-opus-5",
          "name": "Claude Opus 5",
          "providerId": "anthropic",
          "score": 93.9,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gpt-5-6-sol",
          "name": "GPT-5.6 Sol",
          "providerId": "openai",
          "score": 93.5,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "grok-4-5",
          "name": "Grok 4.5",
          "providerId": "xai",
          "score": 93.4,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gpt-5-6-terra",
          "name": "GPT-5.6 Terra",
          "providerId": "openai",
          "score": 93.3,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gpt-5-4",
          "name": "GPT-5.4",
          "providerId": "openai",
          "score": 93.3,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "kimi-k3",
          "name": "Kimi K3",
          "providerId": "moonshot",
          "score": 93.1,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gemini-3-5-flash",
          "name": "Gemini 3.5 Flash",
          "providerId": "google",
          "score": 92.8,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "qwen3-8-max",
          "name": "Qwen3.8-Max",
          "providerId": "qwen",
          "score": 92.7,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "glm-5-2",
          "name": "GLM-5.2",
          "providerId": "zai",
          "score": 91.9,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gpt-5-6-luna",
          "name": "GPT-5.6 Luna",
          "providerId": "openai",
          "score": 91.6,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gpt-5-2",
          "name": "GPT-5.2",
          "providerId": "openai",
          "score": 91.4,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "claude-opus-4-8",
          "name": "Claude Opus 4.8",
          "providerId": "anthropic",
          "score": 91,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "deepseek-v4-pro",
          "name": "DeepSeek V4 Pro",
          "providerId": "deepseek",
          "score": 90.9,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "minimax-m3",
          "name": "MiniMax M3",
          "providerId": "minimax",
          "score": 90.9,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "glm-5-3",
          "name": "GLM-5.3",
          "providerId": "zai",
          "score": 90.9,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "kimi-k2-6",
          "name": "Kimi K2.6",
          "providerId": "moonshot",
          "score": 90.8,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gpt-5-5",
          "name": "GPT-5.5",
          "providerId": "openai",
          "score": 90.7,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "claude-sonnet-5",
          "name": "Claude Sonnet 5",
          "providerId": "anthropic",
          "score": 90.5,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "claude-opus-4-6",
          "name": "Claude Opus 4.6",
          "providerId": "anthropic",
          "score": 90.5,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "claude-opus-4-7",
          "name": "Claude Opus 4.7",
          "providerId": "anthropic",
          "score": 90.2,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "glm-5-3-flash",
          "name": "GLM-5.3-Flash",
          "providerId": "zai",
          "score": 90.2,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "glm-5-1",
          "name": "GLM-5.1",
          "providerId": "zai",
          "score": 89.9,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gemini-3-flash-preview",
          "name": "Gemini 3 Flash Preview",
          "providerId": "google",
          "score": 89.4,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "grok-4-20-reasoning",
          "name": "Grok 4.20 Reasoning",
          "providerId": "xai",
          "score": 89.3,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "grok-4-3",
          "name": "Grok 4.3",
          "providerId": "xai",
          "score": 88.8,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "qwen3-6-plus",
          "name": "Qwen3.6-Plus",
          "providerId": "qwen",
          "score": 88.4,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "kimi-k2-7-code",
          "name": "Kimi K2.7 Code",
          "providerId": "moonshot",
          "score": 87.9,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "qwen3-7-plus",
          "name": "Qwen3.7-Plus",
          "providerId": "qwen",
          "score": 87.9,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "glm-5",
          "name": "GLM-5",
          "providerId": "zai",
          "score": 87.8,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gpt-5-1",
          "name": "GPT-5.1",
          "providerId": "openai",
          "score": 87.6,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "claude-sonnet-4-6",
          "name": "Claude Sonnet 4.6",
          "providerId": "anthropic",
          "score": 87.4,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gpt-5-4-mini",
          "name": "GPT-5.4 mini",
          "providerId": "openai",
          "score": 86.9,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gpt-5",
          "name": "GPT-5",
          "providerId": "openai",
          "score": 86.2,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "qwen3-6-27b",
          "name": "Qwen3.6 27B",
          "providerId": "qwen",
          "score": 85.9,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "qwen3-5-397b-a17b",
          "name": "Qwen3.5 397B-A17B",
          "providerId": "qwen",
          "score": 85.9,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gemini-2-5-pro",
          "name": "Gemini 2.5 Pro",
          "providerId": "google",
          "score": 85.3,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "qwen3-5-plus",
          "name": "Qwen3.5-Plus",
          "providerId": "qwen",
          "score": 84.8,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "qwen3-6-35b-a3b",
          "name": "Qwen3.6 35B-A3B",
          "providerId": "qwen",
          "score": 83.8,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "qwen3-5-35b-a3b",
          "name": "Qwen3.5 35B-A3B",
          "providerId": "qwen",
          "score": 83.5,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gemini-3-5-flash-lite",
          "name": "Gemini 3.5 Flash-Lite",
          "providerId": "google",
          "score": 83.3,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "qwen3-6-flash",
          "name": "Qwen3.6-Flash",
          "providerId": "qwen",
          "score": 83.3,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "glm-4-7",
          "name": "GLM-4.7",
          "providerId": "zai",
          "score": 83.3,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "claude-sonnet-4-5",
          "name": "Claude Sonnet 4.5",
          "providerId": "anthropic",
          "score": 82.3,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "qwen3-7-flash",
          "name": "Qwen3.7-Flash",
          "providerId": "qwen",
          "score": 82.3,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "qwen3-5-flash",
          "name": "Qwen3.5-Flash",
          "providerId": "qwen",
          "score": 82.3,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "o3",
          "name": "o3",
          "providerId": "openai",
          "score": 81.8,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gemini-3-1-flash-lite",
          "name": "Gemini 3.1 Flash-Lite",
          "providerId": "google",
          "score": 81.8,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "claude-sonnet-3-7",
          "name": "Claude 3.7 Sonnet",
          "providerId": "anthropic",
          "score": 79.7,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "claude-sonnet-4",
          "name": "Claude Sonnet 4",
          "providerId": "anthropic",
          "score": 79.2,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gpt-5-4-nano",
          "name": "GPT-5.4 nano",
          "providerId": "openai",
          "score": 78.5,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "claude-opus-4-1",
          "name": "Claude Opus 4.1",
          "providerId": "anthropic",
          "score": 77.3,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "claude-opus-4",
          "name": "Claude Opus 4",
          "providerId": "anthropic",
          "score": 76.3,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gpt-oss-120b",
          "name": "gpt-oss-120b",
          "providerId": "openai",
          "score": 75.8,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gpt-5-mini",
          "name": "GPT-5 mini",
          "providerId": "openai",
          "score": 75,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "qwen3-max",
          "name": "Qwen3-Max",
          "providerId": "qwen",
          "score": 72.6,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "deepseek-r1",
          "name": "DeepSeek R1",
          "providerId": "deepseek",
          "score": 71.7,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "claude-haiku-4-5",
          "name": "Claude Haiku 4.5",
          "providerId": "anthropic",
          "score": 71.2,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "qwen3-235b-a22b",
          "name": "Qwen3 235B-A22B",
          "providerId": "qwen",
          "score": 70.7,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gpt-5-nano",
          "name": "GPT-5 nano",
          "providerId": "openai",
          "score": 69.4,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gpt-4-1",
          "name": "GPT-4.1",
          "providerId": "openai",
          "score": 66.9,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gpt-4-1-mini",
          "name": "GPT-4.1 mini",
          "providerId": "openai",
          "score": 65.8,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gpt-oss-20b",
          "name": "gpt-oss-20b",
          "providerId": "openai",
          "score": 60.8,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "glm-4-7-flash",
          "name": "GLM-4.7-Flash",
          "providerId": "zai",
          "score": 60.5,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "deepseek-v3",
          "name": "DeepSeek V3",
          "providerId": "deepseek",
          "score": 56.5,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "phi-4",
          "name": "Phi-4",
          "providerId": "microsoft",
          "score": 56.1,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "claude-sonnet-3-5",
          "name": "Claude 3.5 Sonnet",
          "providerId": "anthropic",
          "score": 55.3,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gpt-4o",
          "name": "GPT-4o",
          "providerId": "openai",
          "score": 49.2,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gpt-4-1-nano",
          "name": "GPT-4.1 nano",
          "providerId": "openai",
          "score": 48.9,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "llama-3-3-70b-instruct",
          "name": "Llama 3.3 70B Instruct",
          "providerId": "meta",
          "score": 47.4,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        },
        {
          "modelId": "gpt-4o-mini",
          "name": "GPT-4o mini",
          "providerId": "openai",
          "score": 37.7,
          "unit": "percent",
          "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond",
          "measuredAt": "2026-09-17"
        }
      ]
    },
    {
      "benchmark": "GPQA Diamond",
      "count": 70,
      "scores": [
        {
          "modelId": "gpt-6-astra",
          "name": "GPT-6 Astra",
          "providerId": "openai",
          "score": 96,
          "unit": "percent",
          "sourceUrl": "https://openai.com/index/gpt-6-astra/",
          "measuredAt": "2026-09-03"
        },
        {
          "modelId": "gemini-3-8-flash",
          "name": "Gemini 3.8 Flash",
          "providerId": "google",
          "score": 95.3,
          "unit": "percent",
          "sourceUrl": "https://openai.com/index/gpt-6-astra/",
          "measuredAt": "2026-09-03"
        },
        {
          "modelId": "gpt-5-6-sol",
          "name": "GPT-5.6 Sol",
          "providerId": "openai",
          "score": 94.6,
          "unit": "percent",
          "sourceUrl": "https://openai.com/index/gpt-5-6/",
          "measuredAt": "2026-07-09"
        },
        {
          "modelId": "gpt-5-4-pro",
          "name": "GPT-5.4 Pro",
          "providerId": "openai",
          "score": 94.4,
          "unit": "percent",
          "sourceUrl": "https://openai.com/index/introducing-gpt-5-4/",
          "measuredAt": "2026-03-05"
        },
        {
          "modelId": "gemini-3-1-pro-preview",
          "name": "Gemini 3.1 Pro Preview",
          "providerId": "google",
          "score": 94.3,
          "unit": "percent",
          "sourceUrl": "https://openai.com/index/gpt-5-6/",
          "measuredAt": "2026-07-09"
        },
        {
          "modelId": "claude-opus-4-7",
          "name": "Claude Opus 4.7",
          "providerId": "anthropic",
          "score": 94.2,
          "unit": "percent",
          "sourceUrl": "https://www-cdn.anthropic.com/0b4915911bb0d19eca5b5ee635c80fef830a37ea.pdf",
          "measuredAt": "2026-05-28"
        },
        {
          "modelId": "claude-opus-4-8",
          "name": "Claude Opus 4.8",
          "providerId": "anthropic",
          "score": 93.6,
          "unit": "percent",
          "sourceUrl": "https://www-cdn.anthropic.com/0b4915911bb0d19eca5b5ee635c80fef830a37ea.pdf",
          "measuredAt": "2026-05-28"
        },
        {
          "modelId": "kimi-k3",
          "name": "Kimi K3",
          "providerId": "moonshot",
          "score": 93.5,
          "unit": "percent",
          "sourceUrl": "https://github.com/MoonshotAI/Kimi-K3",
          "measuredAt": "2026-07-27"
        },
        {
          "modelId": "gpt-5-2-pro",
          "name": "GPT-5.2 Pro",
          "providerId": "openai",
          "score": 93.2,
          "unit": "percent",
          "sourceUrl": "https://openai.com/index/introducing-gpt-5-2/",
          "measuredAt": "2025-12-11"
        },
        {
          "modelId": "gpt-5-6-terra",
          "name": "GPT-5.6 Terra",
          "providerId": "openai",
          "score": 92.9,
          "unit": "percent",
          "sourceUrl": "https://openai.com/index/gpt-5-6/",
          "measuredAt": "2026-07-09"
        },
        {
          "modelId": "gpt-5-4",
          "name": "GPT-5.4",
          "providerId": "openai",
          "score": 92.8,
          "unit": "percent",
          "sourceUrl": "https://openai.com/index/introducing-gpt-5-4/",
          "measuredAt": "2026-03-05"
        },
        {
          "modelId": "qwen3-8-max",
          "name": "Qwen3.8-Max",
          "providerId": "qwen",
          "score": 92.6,
          "unit": "percent",
          "sourceUrl": "https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B",
          "measuredAt": "2026-08-12"
        },
        {
          "modelId": "qwen3-8-2-4t-a95b",
          "name": "Qwen3.8 2.4T-A95B",
          "providerId": "qwen",
          "score": 92.6,
          "unit": "percent",
          "sourceUrl": "https://docs.modelstudio.console.alibabacloud.com/en/model-studio/qwen3-8-2-4t-a95b",
          "measuredAt": "2026-08-12"
        },
        {
          "modelId": "gpt-5-2",
          "name": "GPT-5.2",
          "providerId": "openai",
          "score": 92.4,
          "unit": "percent",
          "sourceUrl": "https://openai.com/index/introducing-gpt-5-2/",
          "measuredAt": "2025-12-11"
        },
        {
          "modelId": "dola-seed-2-0-pro",
          "name": "Dola Seed 2.0 Pro",
          "providerId": "bytedance",
          "score": 92.4,
          "unit": "percent",
          "sourceUrl": "https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf",
          "measuredAt": "2026-02-16"
        },
        {
          "modelId": "gpt-5-6-luna",
          "name": "GPT-5.6 Luna",
          "providerId": "openai",
          "score": 92.3,
          "unit": "percent",
          "sourceUrl": "https://openai.com/index/gpt-5-6/",
          "measuredAt": "2026-07-09"
        },
        {
          "modelId": "qwen3-8-flash-next",
          "name": "Qwen3.8-Flash-Next",
          "providerId": "qwen",
          "score": 91.7,
          "unit": "percent",
          "sourceUrl": "https://huggingface.co/Qwen/Qwen3.8-Flash-Next",
          "measuredAt": "2026-09-16"
        },
        {
          "modelId": "claude-opus-4-6",
          "name": "Claude Opus 4.6",
          "providerId": "anthropic",
          "score": 91.3,
          "unit": "percent",
          "sourceUrl": "https://www-cdn.anthropic.com/14e4fb01875d2a69f646fa5e574dea2b1c0ff7b5.pdf",
          "measuredAt": "2026-02-05"
        },
        {
          "modelId": "deepseek-v4-1-flash",
          "name": "DeepSeek V4.1 Flash",
          "providerId": "deepseek",
          "score": 90.9,
          "unit": "percent",
          "sourceUrl": "https://api-docs.deepseek.com/updates/",
          "measuredAt": "2026-09-10"
        },
        {
          "modelId": "hy3",
          "name": "Hy3",
          "providerId": "tencent",
          "score": 90.4,
          "unit": "percent",
          "sourceUrl": "https://github.com/Tencent-Hunyuan/Hy3",
          "measuredAt": "2026-07-06"
        },
        {
          "modelId": "gemini-3-flash-preview",
          "name": "Gemini 3 Flash Preview",
          "providerId": "google",
          "score": 90.4,
          "unit": "percent",
          "sourceUrl": "https://blog.google/products-and-platforms/products/gemini/gemini-3-flash/",
          "measuredAt": "2025-12-17"
        },
        {
          "modelId": "claude-sonnet-4-6",
          "name": "Claude Sonnet 4.6",
          "providerId": "anthropic",
          "score": 89.9,
          "unit": "percent",
          "sourceUrl": "https://www-cdn.anthropic.com/78073f739564e986ff3e28522761a7a0b4484f84.pdf",
          "measuredAt": "2026-02-17"
        },
        {
          "modelId": "qwen3-8-27b",
          "name": "Qwen3.8 27B",
          "providerId": "qwen",
          "score": 89.2,
          "unit": "percent",
          "sourceUrl": "https://huggingface.co/Qwen/Qwen3.8-27B",
          "measuredAt": "2026-08-14"
        },
        {
          "modelId": "longcat-2-0",
          "name": "LongCat 2.0",
          "providerId": "meituan",
          "score": 88.9,
          "unit": "percent",
          "sourceUrl": "https://github.com/meituan-longcat/LongCat-2.0",
          "measuredAt": "2026-06-30"
        },
        {
          "modelId": "gpt-5-pro",
          "name": "GPT-5 Pro",
          "providerId": "openai",
          "score": 88.4,
          "unit": "percent",
          "sourceUrl": "https://openai.com/index/introducing-gpt-5/",
          "measuredAt": "2025-08-07"
        },
        {
          "modelId": "qwen3-5-397b-a17b",
          "name": "Qwen3.5 397B-A17B",
          "providerId": "qwen",
          "score": 88.4,
          "unit": "percent",
          "sourceUrl": "https://qwen.ai/blog?id=qwen3.6-27b",
          "measuredAt": "2026-04-21"
        },
        {
          "modelId": "gpt-5-4-mini",
          "name": "GPT-5.4 mini",
          "providerId": "openai",
          "score": 88,
          "unit": "percent",
          "sourceUrl": "https://openai.com/index/introducing-gpt-5-4-mini-and-nano/",
          "measuredAt": "2026-03-17"
        },
        {
          "modelId": "nvidia-nemotron-3-ultra-550b-a55b-nvfp4",
          "name": "NVIDIA Nemotron 3 Ultra 550B-A55B NVFP4",
          "providerId": "nvidia",
          "score": 87.9,
          "unit": "percent",
          "sourceUrl": "https://build.nvidia.com/nvidia/nemotron-3-ultra-550b-a55b/modelcard",
          "measuredAt": "2026-06-04"
        },
        {
          "modelId": "qwen3-6-27b",
          "name": "Qwen3.6 27B",
          "providerId": "qwen",
          "score": 87.8,
          "unit": "percent",
          "sourceUrl": "https://qwen.ai/blog?id=qwen3.6-27b",
          "measuredAt": "2026-04-21"
        },
        {
          "modelId": "gemini-3-1-flash-lite",
          "name": "Gemini 3.1 Flash-Lite",
          "providerId": "google",
          "score": 86.9,
          "unit": "percent",
          "sourceUrl": "https://deepmind.google/models/model-cards/gemini-3-1-flash-lite/",
          "measuredAt": "2026-03-03"
        },
        {
          "modelId": "gemini-2-5-pro",
          "name": "Gemini 2.5 Pro",
          "providerId": "google",
          "score": 86.4,
          "unit": "percent",
          "sourceUrl": "https://storage.googleapis.com/deepmind-media/gemini/gemini_v2_5_report.pdf",
          "measuredAt": "2025-06-17"
        },
        {
          "modelId": "ernie-5-0",
          "name": "ERNIE 5.0",
          "providerId": "baidu",
          "score": 86.36,
          "unit": "percent",
          "sourceUrl": "https://ernie.baidu.com/blog/posts/ernie5.0/",
          "measuredAt": "2026-02-06"
        },
        {
          "modelId": "solar-open-2-250b",
          "name": "Solar Open 2 250B-A15B",
          "providerId": "upstage",
          "score": 86.3,
          "unit": "percent",
          "sourceUrl": "https://huggingface.co/upstage/Solar-Open2-250B",
          "measuredAt": "2026-07-24"
        },
        {
          "modelId": "qwen3-6-35b-a3b",
          "name": "Qwen3.6 35B-A3B",
          "providerId": "qwen",
          "score": 86,
          "unit": "percent",
          "sourceUrl": "https://qwen.ai/blog?id=qwen3.6-27b",
          "measuredAt": "2026-04-21"
        },
        {
          "modelId": "grok-4-fast-reasoning",
          "name": "Grok 4 Fast Reasoning",
          "providerId": "xai",
          "score": 85.7,
          "unit": "percent",
          "sourceUrl": "https://x.ai/news/grok-4-fast",
          "measuredAt": "2025-09-19"
        },
        {
          "modelId": "qwen3-5-27b",
          "name": "Qwen3.5 27B",
          "providerId": "qwen",
          "score": 85.5,
          "unit": "percent",
          "sourceUrl": "https://qwen.ai/blog?id=qwen3.6-27b",
          "measuredAt": "2026-04-21"
        },
        {
          "modelId": "gemma-4-31b",
          "name": "Gemma 4 31B",
          "providerId": "google",
          "score": 84.3,
          "unit": "percent",
          "sourceUrl": "https://ai.google.dev/gemma/docs/core/model_card_4",
          "measuredAt": "2026-04-02"
        },
        {
          "modelId": "mai-thinking-1",
          "name": "MAI-Thinking-1",
          "providerId": "microsoft",
          "score": 84.2,
          "unit": "percent",
          "sourceUrl": "https://microsoft.ai/pdf/mai-thinking-1.pdf",
          "measuredAt": "2026-08-12"
        },
        {
          "modelId": "claude-sonnet-4-5",
          "name": "Claude Sonnet 4.5",
          "providerId": "anthropic",
          "score": 83.4,
          "unit": "percent",
          "sourceUrl": "https://www-cdn.anthropic.com/78073f739564e986ff3e28522761a7a0b4484f84.pdf",
          "measuredAt": "2026-02-17"
        },
        {
          "modelId": "gpt-5-4-nano",
          "name": "GPT-5.4 nano",
          "providerId": "openai",
          "score": 82.8,
          "unit": "percent",
          "sourceUrl": "https://openai.com/index/introducing-gpt-5-4-mini-and-nano/",
          "measuredAt": "2026-03-17"
        },
        {
          "modelId": "gemini-2-5-flash",
          "name": "Gemini 2.5 Flash",
          "providerId": "google",
          "score": 82.8,
          "unit": "percent",
          "sourceUrl": "https://storage.googleapis.com/deepmind-media/gemini/gemini_v2_5_report.pdf",
          "measuredAt": "2025-06-17"
        },
        {
          "modelId": "deepseek-v3-2",
          "name": "DeepSeek V3.2",
          "providerId": "deepseek",
          "score": 82.4,
          "unit": "percent",
          "sourceUrl": "https://modelscope.cn/models/deepseek-ai/DeepSeek-V3.2/resolve/master/assets/paper.pdf",
          "measuredAt": "2025-12-01"
        },
        {
          "modelId": "gemma-4-26b-a4b",
          "name": "Gemma 4 26B A4B",
          "providerId": "google",
          "score": 82.3,
          "unit": "percent",
          "sourceUrl": "https://ai.google.dev/gemma/docs/core/model_card_4",
          "measuredAt": "2026-04-02"
        },
        {
          "modelId": "gpt-5-mini",
          "name": "GPT-5 mini",
          "providerId": "openai",
          "score": 81.6,
          "unit": "percent",
          "sourceUrl": "https://openai.com/index/introducing-gpt-5-4-mini-and-nano/",
          "measuredAt": "2026-03-17"
        },
        {
          "modelId": "gpt-oss-120b",
          "name": "gpt-oss-120b",
          "providerId": "openai",
          "score": 80.1,
          "unit": "percent",
          "sourceUrl": "https://cdn.openai.com/pdf/419b6906-9da6-406c-a19d-1bb078ac7637/oai_gpt-oss_model_card.pdf",
          "measuredAt": "2025-08-05"
        },
        {
          "modelId": "amazon-nova-2-lite",
          "name": "Amazon Nova 2 Lite",
          "providerId": "amazon",
          "score": 79.6,
          "unit": "percent",
          "sourceUrl": "https://cdn.amazon.science/c5/3d/84514a224666b5be6de4b43ef4aa/nova-2-0-technical-report2.pdf",
          "measuredAt": "2025-12-02"
        },
        {
          "modelId": "gemma-4-12b",
          "name": "Gemma 4 12B Unified",
          "providerId": "google",
          "score": 78.8,
          "unit": "percent",
          "sourceUrl": "https://ai.google.dev/gemma/docs/core/model_card_4",
          "measuredAt": "2026-06-03"
        },
        {
          "modelId": "nvidia-nemotron-3-5-lightning-30b-a3b-nvfp4",
          "name": "NVIDIA Nemotron 3.5 Lightning 30B-A3B NVFP4",
          "providerId": "nvidia",
          "score": 75.57,
          "unit": "percent",
          "sourceUrl": "https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-NVFP4",
          "measuredAt": "2026-08-11"
        },
        {
          "modelId": "longcat-flash-chat",
          "name": "LongCat Flash Chat",
          "providerId": "meituan",
          "score": 73.23,
          "unit": "percent",
          "sourceUrl": "https://github.com/meituan-longcat/LongCat-Flash-Chat",
          "measuredAt": "2025-09-01"
        },
        {
          "modelId": "gpt-oss-20b",
          "name": "gpt-oss-20b",
          "providerId": "openai",
          "score": 71.5,
          "unit": "percent",
          "sourceUrl": "https://cdn.openai.com/pdf/419b6906-9da6-406c-a19d-1bb078ac7637/oai_gpt-oss_model_card.pdf",
          "measuredAt": "2025-08-05"
        },
        {
          "modelId": "deepseek-r1",
          "name": "DeepSeek R1",
          "providerId": "deepseek",
          "score": 71.5,
          "unit": "percent",
          "sourceUrl": "https://github.com/deepseek-ai/DeepSeek-R1/blob/main/README.md",
          "measuredAt": "2025-01-20"
        },
        {
          "modelId": "mistral-small-4",
          "name": "Mistral Small 4",
          "providerId": "mistral",
          "score": 71.2,
          "unit": "percent",
          "sourceUrl": "https://huggingface.co/mistralai/Mistral-Small-4-119B-2603",
          "measuredAt": "2026-03-16"
        },
        {
          "modelId": "ministral-3-14b",
          "name": "Ministral 3 14B",
          "providerId": "mistral",
          "score": 71.2,
          "unit": "percent",
          "sourceUrl": "https://huggingface.co/mistralai/Ministral-3-14B-Instruct-2512-BF16",
          "measuredAt": "2025-12-02"
        },
        {
          "modelId": "llama-4-maverick",
          "name": "Llama 4 Maverick",
          "providerId": "meta",
          "score": 69.8,
          "unit": "percent",
          "sourceUrl": "https://huggingface.co/meta-llama/Llama-4-Maverick-17B-128E-Instruct",
          "measuredAt": "2025-04-05"
        },
        {
          "modelId": "ministral-3-8b",
          "name": "Ministral 3 8B",
          "providerId": "mistral",
          "score": 66.8,
          "unit": "percent",
          "sourceUrl": "https://huggingface.co/mistralai/Ministral-3-8B-Instruct-2512-BF16",
          "measuredAt": "2025-12-02"
        },
        {
          "modelId": "gemini-2-5-flash-lite",
          "name": "Gemini 2.5 Flash-Lite",
          "providerId": "google",
          "score": 66.7,
          "unit": "percent",
          "sourceUrl": "https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-5-Flash-Lite-Model-Card.pdf",
          "measuredAt": "2025-06-17"
        },
        {
          "modelId": "gpt-4-1",
          "name": "GPT-4.1",
          "providerId": "openai",
          "score": 66.3,
          "unit": "percent",
          "sourceUrl": "https://openai.com/index/gpt-4-1/",
          "measuredAt": "2025-04-14"
        },
        {
          "modelId": "phi-4-reasoning",
          "name": "Phi-4 Reasoning",
          "providerId": "microsoft",
          "score": 65.8,
          "unit": "percent",
          "sourceUrl": "https://huggingface.co/microsoft/Phi-4-reasoning",
          "measuredAt": "2025-04-30"
        },
        {
          "modelId": "gpt-4-1-mini",
          "name": "GPT-4.1 mini",
          "providerId": "openai",
          "score": 65,
          "unit": "percent",
          "sourceUrl": "https://openai.com/index/gpt-4-1/",
          "measuredAt": "2025-04-14"
        },
        {
          "modelId": "deepseek-v3",
          "name": "DeepSeek V3",
          "providerId": "deepseek",
          "score": 59.1,
          "unit": "percent",
          "sourceUrl": "https://github.com/deepseek-ai/DeepSeek-V3",
          "measuredAt": "2024-12-26"
        },
        {
          "modelId": "gemma-4-e4b",
          "name": "Gemma 4 E4B",
          "providerId": "google",
          "score": 58.6,
          "unit": "percent",
          "sourceUrl": "https://ai.google.dev/gemma/docs/core/model_card_4",
          "measuredAt": "2026-04-02"
        },
        {
          "modelId": "llama-4-scout",
          "name": "Llama 4 Scout",
          "providerId": "meta",
          "score": 57.2,
          "unit": "percent",
          "sourceUrl": "https://huggingface.co/meta-llama/Llama-4-Maverick-17B-128E-Instruct",
          "measuredAt": "2025-04-05"
        },
        {
          "modelId": "phi-4",
          "name": "Phi-4",
          "providerId": "microsoft",
          "score": 56.1,
          "unit": "percent",
          "sourceUrl": "https://www.microsoft.com/en-us/research/wp-content/uploads/2024/12/P4TechReport.pdf",
          "measuredAt": "2024-12-12"
        },
        {
          "modelId": "ministral-3-3b",
          "name": "Ministral 3 3B",
          "providerId": "mistral",
          "score": 53.4,
          "unit": "percent",
          "sourceUrl": "https://huggingface.co/mistralai/Ministral-3-3B-Instruct-2512",
          "measuredAt": "2025-12-02"
        },
        {
          "modelId": "phi-4-mini-reasoning",
          "name": "Phi-4 Mini Reasoning",
          "providerId": "microsoft",
          "score": 52,
          "unit": "percent",
          "sourceUrl": "https://huggingface.co/microsoft/Phi-4-mini-reasoning",
          "measuredAt": "2025-04-30"
        },
        {
          "modelId": "llama-3-3-70b-instruct",
          "name": "Llama 3.3 70B Instruct",
          "providerId": "meta",
          "score": 50.5,
          "unit": "percent",
          "sourceUrl": "https://huggingface.co/meta-llama/Llama-4-Maverick-17B-128E-Instruct",
          "measuredAt": "2025-04-05"
        },
        {
          "modelId": "gpt-4-1-nano",
          "name": "GPT-4.1 nano",
          "providerId": "openai",
          "score": 50.3,
          "unit": "percent",
          "sourceUrl": "https://openai.com/index/gpt-4-1/",
          "measuredAt": "2025-04-14"
        },
        {
          "modelId": "gpt-4o",
          "name": "GPT-4o",
          "providerId": "openai",
          "score": 46,
          "unit": "percent",
          "sourceUrl": "https://openai.com/index/gpt-4-1/",
          "measuredAt": "2025-04-14"
        },
        {
          "modelId": "gemma-4-e2b",
          "name": "Gemma 4 E2B",
          "providerId": "google",
          "score": 43.4,
          "unit": "percent",
          "sourceUrl": "https://ai.google.dev/gemma/docs/core/model_card_4",
          "measuredAt": "2026-04-02"
        },
        {
          "modelId": "gpt-4o-mini",
          "name": "GPT-4o mini",
          "providerId": "openai",
          "score": 40.2,
          "unit": "percent",
          "sourceUrl": "https://openai.com/index/gpt-4-1/",
          "measuredAt": "2025-04-14"
        }
      ]
    },
    {
      "benchmark": "GPQA",
      "count": 4,
      "scores": [
        {
          "modelId": "command-a",
          "name": "Command A",
          "providerId": "cohere",
          "score": 50.8,
          "unit": "percent",
          "sourceUrl": "https://cohere.com/research/papers/command-a-technical-report.pdf",
          "measuredAt": "2025-03-13"
        },
        {
          "modelId": "amazon-nova-pro",
          "name": "Amazon Nova Pro",
          "providerId": "amazon",
          "score": 46.9,
          "unit": "percent",
          "sourceUrl": "https://assets.amazon.science/10/0a/0b61d39a4e9aaec16f71ad3d9168/the-amazon-nova-family-of-models-technical-report-and-model-card2-26.pdf",
          "measuredAt": "2024-12-03"
        },
        {
          "modelId": "amazon-nova-lite",
          "name": "Amazon Nova Lite",
          "providerId": "amazon",
          "score": 42,
          "unit": "percent",
          "sourceUrl": "https://assets.amazon.science/10/0a/0b61d39a4e9aaec16f71ad3d9168/the-amazon-nova-family-of-models-technical-report-and-model-card2-26.pdf",
          "measuredAt": "2024-12-03"
        },
        {
          "modelId": "command-r7b",
          "name": "Command R7B",
          "providerId": "cohere",
          "score": 26.3,
          "unit": "percent",
          "sourceUrl": "https://cohere.com/research/papers/command-a-technical-report.pdf",
          "measuredAt": "2025-03-13"
        }
      ]
    },
    {
      "benchmark": "SuperGPQA",
      "count": 2,
      "scores": [
        {
          "modelId": "seed-2-1-pro",
          "name": "Seed2.1 Pro",
          "providerId": "bytedance",
          "score": 70.8,
          "unit": "percent",
          "sourceUrl": "https://seed.bytedance.com/en/seed2_1",
          "measuredAt": "2026-07-13"
        },
        {
          "modelId": "dola-seed-2-1-turbo",
          "name": "Dola Seed 2.1 Turbo",
          "providerId": "bytedance",
          "score": 67.4,
          "unit": "percent",
          "sourceUrl": "https://seed.bytedance.com/en/seed2_1",
          "measuredAt": "2026-07-13"
        }
      ]
    },
    {
      "benchmark": "GPQA (no tools)",
      "count": 2,
      "scores": [
        {
          "modelId": "nvidia-nemotron-3-super-120b-a12b",
          "name": "NVIDIA Nemotron 3 Super 120B-A12B",
          "providerId": "nvidia",
          "score": 79.23,
          "unit": "percent",
          "sourceUrl": "https://build.nvidia.com/nvidia/nemotron-3-super-120b-a12b/modelcard",
          "measuredAt": "2026-03-11"
        },
        {
          "modelId": "nvidia-nemotron-3-nano-30b-a3b-bf16",
          "name": "NVIDIA Nemotron 3 Nano 30B-A3B BF16",
          "providerId": "nvidia",
          "score": 73,
          "unit": "percent",
          "sourceUrl": "https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16",
          "measuredAt": "2025-12-15"
        }
      ]
    },
    {
      "benchmark": "GPQA (with tools)",
      "count": 2,
      "scores": [
        {
          "modelId": "nvidia-nemotron-3-super-120b-a12b",
          "name": "NVIDIA Nemotron 3 Super 120B-A12B",
          "providerId": "nvidia",
          "score": 82.7,
          "unit": "percent",
          "sourceUrl": "https://build.nvidia.com/nvidia/nemotron-3-super-120b-a12b/modelcard",
          "measuredAt": "2026-03-11"
        },
        {
          "modelId": "nvidia-nemotron-3-nano-30b-a3b-bf16",
          "name": "NVIDIA Nemotron 3 Nano 30B-A3B BF16",
          "providerId": "nvidia",
          "score": 75,
          "unit": "percent",
          "sourceUrl": "https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16",
          "measuredAt": "2025-12-15"
        }
      ]
    }
  ],
  "meta": {
    "source": "https://llmmetric.com",
    "attribution": "Please cite llmmetric.com. Upstream records remain subject to their source terms.",
    "generatedAt": "2026-09-18T03:45:43.545Z"
  }
}