{ "data": { "version": "beta-1", "methodologyUrl": "/methodology#llmmetric-index", "minPeers": 3, "activeAxes": [ { "id": "swe-verified", "name": "SWE-bench Verified", "category": "Coding", "peers": 48 }, { "id": "swe-pro", "name": "SWE-bench Pro", "category": "Coding", "peers": 22 }, { "id": "deepswe-1-1", "name": "DeepSWE v1.1", "category": "Coding", "peers": 16 }, { "id": "terminal-science-0-1", "name": "Terminal-Bench Science 0.1", "category": "Coding", "peers": 4 }, { "id": "terminal-4", "name": "Terminal-Bench 4.0", "category": "Coding", "peers": 3 }, { "id": "livecodebench-6", "name": "LiveCodeBench v6", "category": "Coding", "peers": 11 }, { "id": "gpqa", "name": "GPQA Diamond", "category": "Reasoning", "peers": 111 }, { "id": "mmlu-pro", "name": "MMLU-Pro", "category": "Reasoning", "peers": 29 }, { "id": "aime-2025", "name": "AIME 2025", "category": "Reasoning", "peers": 40 }, { "id": "hle-no-tools", "name": "Humanity's Last Exam (no tools)", "category": "Reasoning", "peers": 29 }, { "id": "hle-with-tools", "name": "Humanity's Last Exam (with tools)", "category": "Reasoning", "peers": 14 }, { "id": "frontiermath-1-3-epoch", "name": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "peers": 59 }, { "id": "automation-1-0-6", "name": "AutomationBench 1.0.6", "category": "Agents", "peers": 3 }, { "id": "browsecomp", "name": "BrowseComp", "category": "Agents", "peers": 15 }, { "id": "mcp-atlas", "name": "MCP Atlas", "category": "Agents", "peers": 9 }, { "id": "tau2-telecom", "name": "Tau2-bench Telecom", "category": "Agents", "peers": 8 }, { "id": "toolathlon", "name": "Toolathlon", "category": "Agents", "peers": 6 }, { "id": "osworld-verified", "name": "OSWorld-Verified", "category": "Agents", "peers": 12 } ], "ranked": [ { "modelId": "gpt-6-astra", "name": "GPT-6 Astra", "providerId": "openai", "releaseDate": "2026-09-03", "rank": 1, "score": 90.1, "categoryScores": { "Coding": 87.5, "Reasoning": 99.4, "Agents": 83.3 }, "coverage": 4, "categoryCoverage": 3, "missingCategories": [], "components": [ { "axisId": "terminal-science-0-1", "benchmark": "Terminal-Bench Science 0.1", "category": "Coding", "rawScore": 64.6, "percentile": 87.5, "peerCount": 4, "sourceUrl": "https://openai.com/index/gpt-6-astra/", "measuredAt": "2026-09-03" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 96, "percentile": 99.5, "peerCount": 111, "sourceUrl": "https://openai.com/index/gpt-6-astra/", "measuredAt": "2026-09-03" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 93.7, "percentile": 99.2, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" }, { "axisId": "automation-1-0-6", "benchmark": "AutomationBench 1.0.6", "category": "Agents", "rawScore": 41.4, "percentile": 83.3, "peerCount": 3, "sourceUrl": "https://zapier.com/benchmarks", "measuredAt": "2026-09-03" } ] }, { "modelId": "claude-opus-4-8", "name": "Claude Opus 4.8", "providerId": "anthropic", "releaseDate": "2026-05-28", "rank": 2, "score": 88.6, "categoryScores": { "Coding": 98.4, "Reasoning": 85.2, "Agents": 82.1 }, "coverage": 8, "categoryCoverage": 3, "missingCategories": [], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 88.6, "percentile": 99, "peerCount": 48, "sourceUrl": "https://www-cdn.anthropic.com/0b4915911bb0d19eca5b5ee635c80fef830a37ea.pdf", "measuredAt": "2026-05-28" }, { "axisId": "swe-pro", "benchmark": "SWE-bench Pro", "category": "Coding", "rawScore": 69.2, "percentile": 97.7, "peerCount": 22, "sourceUrl": "https://www-cdn.anthropic.com/0b4915911bb0d19eca5b5ee635c80fef830a37ea.pdf", "measuredAt": "2026-05-28" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 93.6, "percentile": 90.5, "peerCount": 111, "sourceUrl": "https://www-cdn.anthropic.com/0b4915911bb0d19eca5b5ee635c80fef830a37ea.pdf", "measuredAt": "2026-05-28" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 49.8, "percentile": 91.4, "peerCount": 29, "sourceUrl": "https://www-cdn.anthropic.com/0b4915911bb0d19eca5b5ee635c80fef830a37ea.pdf", "measuredAt": "2026-05-28" }, { "axisId": "hle-with-tools", "benchmark": "Humanity's Last Exam (with tools)", "category": "Reasoning", "rawScore": 57.9, "percentile": 75, "peerCount": 14, "sourceUrl": "https://www-cdn.anthropic.com/0b4915911bb0d19eca5b5ee635c80fef830a37ea.pdf", "measuredAt": "2026-05-28" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 80, "percentile": 83.9, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" }, { "axisId": "browsecomp", "benchmark": "BrowseComp", "category": "Agents", "rawScore": 84.3, "percentile": 76.7, "peerCount": 15, "sourceUrl": "https://www-cdn.anthropic.com/0b4915911bb0d19eca5b5ee635c80fef830a37ea.pdf", "measuredAt": "2026-05-28" }, { "axisId": "osworld-verified", "benchmark": "OSWorld-Verified", "category": "Agents", "rawScore": 83.4, "percentile": 87.5, "peerCount": 12, "sourceUrl": "https://www-cdn.anthropic.com/0b4915911bb0d19eca5b5ee635c80fef830a37ea.pdf", "measuredAt": "2026-05-28" } ] }, { "modelId": "kimi-k3", "name": "Kimi K3", "providerId": "moonshot", "releaseDate": "2026-07-16", "rank": 3, "score": 77.1, "categoryScores": { "Coding": 59.4, "Reasoning": 76.2, "Agents": 95.8 }, "coverage": 6, "categoryCoverage": 3, "missingCategories": [], "components": [ { "axisId": "deepswe-1-1", "benchmark": "DeepSWE v1.1", "category": "Coding", "rawScore": 67.3, "percentile": 59.4, "peerCount": 16, "sourceUrl": "https://www.kimi.com/en/blog/kimi-k3", "measuredAt": "2026-07-16" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 93.5, "percentile": 89.6, "peerCount": 111, "sourceUrl": "https://github.com/MoonshotAI/Kimi-K3", "measuredAt": "2026-07-27" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 43.5, "percentile": 84.5, "peerCount": 29, "sourceUrl": "https://github.com/MoonshotAI/Kimi-K3", "measuredAt": "2026-07-27" }, { "axisId": "hle-with-tools", "benchmark": "Humanity's Last Exam (with tools)", "category": "Reasoning", "rawScore": 56, "percentile": 53.6, "peerCount": 14, "sourceUrl": "https://github.com/MoonshotAI/Kimi-K3", "measuredAt": "2026-07-27" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 72.2, "percentile": 77.1, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" }, { "axisId": "osworld-verified", "benchmark": "OSWorld-Verified", "category": "Agents", "rawScore": 84.8, "percentile": 95.8, "peerCount": 12, "sourceUrl": "https://github.com/MoonshotAI/Kimi-K3", "measuredAt": "2026-07-27" } ] }, { "modelId": "claude-opus-4-7", "name": "Claude Opus 4.7", "providerId": "anthropic", "releaseDate": "2026-04-16", "rank": 4, "score": 75.5, "categoryScores": { "Coding": 90.5, "Reasoning": 75.5, "Agents": 60.4 }, "coverage": 8, "categoryCoverage": 3, "missingCategories": [], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 87.6, "percentile": 96.9, "peerCount": 48, "sourceUrl": "https://www-cdn.anthropic.com/0b4915911bb0d19eca5b5ee635c80fef830a37ea.pdf", "measuredAt": "2026-05-28" }, { "axisId": "swe-pro", "benchmark": "SWE-bench Pro", "category": "Coding", "rawScore": 64.3, "percentile": 84.1, "peerCount": 22, "sourceUrl": "https://www-cdn.anthropic.com/0b4915911bb0d19eca5b5ee635c80fef830a37ea.pdf", "measuredAt": "2026-05-28" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 94.2, "percentile": 94.1, "peerCount": 111, "sourceUrl": "https://www-cdn.anthropic.com/0b4915911bb0d19eca5b5ee635c80fef830a37ea.pdf", "measuredAt": "2026-05-28" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 46.9, "percentile": 87.9, "peerCount": 29, "sourceUrl": "https://www-cdn.anthropic.com/0b4915911bb0d19eca5b5ee635c80fef830a37ea.pdf", "measuredAt": "2026-05-28" }, { "axisId": "hle-with-tools", "benchmark": "Humanity's Last Exam (with tools)", "category": "Reasoning", "rawScore": 54.7, "percentile": 46.4, "peerCount": 14, "sourceUrl": "https://www-cdn.anthropic.com/0b4915911bb0d19eca5b5ee635c80fef830a37ea.pdf", "measuredAt": "2026-05-28" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 70.2, "percentile": 73.7, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" }, { "axisId": "browsecomp", "benchmark": "BrowseComp", "category": "Agents", "rawScore": 79.8, "percentile": 50, "peerCount": 15, "sourceUrl": "https://www-cdn.anthropic.com/0b4915911bb0d19eca5b5ee635c80fef830a37ea.pdf", "measuredAt": "2026-05-28" }, { "axisId": "osworld-verified", "benchmark": "OSWorld-Verified", "category": "Agents", "rawScore": 82.8, "percentile": 70.8, "peerCount": 12, "sourceUrl": "https://www-cdn.anthropic.com/0b4915911bb0d19eca5b5ee635c80fef830a37ea.pdf", "measuredAt": "2026-05-28" } ] }, { "modelId": "claude-opus-4-6", "name": "Claude Opus 4.6", "providerId": "anthropic", "releaseDate": "2026-02-05", "rank": 5, "score": 73.8, "categoryScores": { "Coding": 92.7, "Reasoning": 72, "Agents": 56.7 }, "coverage": 6, "categoryCoverage": 3, "missingCategories": [], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 80.8, "percentile": 92.7, "peerCount": 48, "sourceUrl": "https://www-cdn.anthropic.com/14e4fb01875d2a69f646fa5e574dea2b1c0ff7b5.pdf", "measuredAt": "2026-02-05" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 91.3, "percentile": 77.9, "peerCount": 111, "sourceUrl": "https://www-cdn.anthropic.com/14e4fb01875d2a69f646fa5e574dea2b1c0ff7b5.pdf", "measuredAt": "2026-02-05" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 66, "percentile": 66.1, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" }, { "axisId": "mcp-atlas", "benchmark": "MCP Atlas", "category": "Agents", "rawScore": 59.5, "percentile": 38.9, "peerCount": 9, "sourceUrl": "https://www-cdn.anthropic.com/14e4fb01875d2a69f646fa5e574dea2b1c0ff7b5.pdf", "measuredAt": "2026-02-05" }, { "axisId": "tau2-telecom", "benchmark": "Tau2-bench Telecom", "category": "Agents", "rawScore": 99.3, "percentile": 93.8, "peerCount": 8, "sourceUrl": "https://www-cdn.anthropic.com/14e4fb01875d2a69f646fa5e574dea2b1c0ff7b5.pdf", "measuredAt": "2026-02-05" }, { "axisId": "osworld-verified", "benchmark": "OSWorld-Verified", "category": "Agents", "rawScore": 72.7, "percentile": 37.5, "peerCount": 12, "sourceUrl": "https://www-cdn.anthropic.com/14e4fb01875d2a69f646fa5e574dea2b1c0ff7b5.pdf", "measuredAt": "2026-02-05" } ] }, { "modelId": "claude-opus-5-5", "name": "Claude Opus 5.5", "providerId": "anthropic", "releaseDate": "2026-09-22", "rank": 6, "score": 73.1, "categoryScores": { "Coding": 72.9, "Reasoning": 96.4, "Agents": 50 }, "coverage": 4, "categoryCoverage": 3, "missingCategories": [], "components": [ { "axisId": "terminal-science-0-1", "benchmark": "Terminal-Bench Science 0.1", "category": "Coding", "rawScore": 58.7, "percentile": 62.5, "peerCount": 4, "sourceUrl": "https://www.anthropic.com/claude-opus-5-5", "measuredAt": "2026-09-22" }, { "axisId": "terminal-4", "benchmark": "Terminal-Bench 4.0", "category": "Coding", "rawScore": 66.4, "percentile": 83.3, "peerCount": 3, "sourceUrl": "https://www.anthropic.com/claude-opus-5-5", "measuredAt": "2026-09-22" }, { "axisId": "hle-with-tools", "benchmark": "Humanity's Last Exam (with tools)", "category": "Reasoning", "rawScore": 67.7, "percentile": 96.4, "peerCount": 14, "sourceUrl": "https://www.anthropic.com/claude-opus-5-5", "measuredAt": "2026-09-22" }, { "axisId": "automation-1-0-6", "benchmark": "AutomationBench 1.0.6", "category": "Agents", "rawScore": 40, "percentile": 50, "peerCount": 3, "sourceUrl": "https://zapier.com/benchmarks", "measuredAt": "2026-09-22" } ] }, { "modelId": "gpt-5-5", "name": "GPT-5.5", "providerId": "openai", "releaseDate": "2026-04-23", "rank": 7, "score": 69.2, "categoryScores": { "Coding": 52.3, "Reasoning": 80.8, "Agents": 74.4 }, "coverage": 8, "categoryCoverage": 3, "missingCategories": [], "components": [ { "axisId": "swe-pro", "benchmark": "SWE-bench Pro", "category": "Coding", "rawScore": 58.6, "percentile": 52.3, "peerCount": 22, "sourceUrl": "https://openai.com/index/introducing-gpt-5-5/", "measuredAt": "2026-04-23" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 90.7, "percentile": 72.5, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 85.3, "percentile": 89, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" }, { "axisId": "browsecomp", "benchmark": "BrowseComp", "category": "Agents", "rawScore": 84.4, "percentile": 83.3, "peerCount": 15, "sourceUrl": "https://openai.com/index/introducing-gpt-5-5/", "measuredAt": "2026-04-23" }, { "axisId": "mcp-atlas", "benchmark": "MCP Atlas", "category": "Agents", "rawScore": 75.3, "percentile": 72.2, "peerCount": 9, "sourceUrl": "https://openai.com/index/introducing-gpt-5-5/", "measuredAt": "2026-04-23" }, { "axisId": "tau2-telecom", "benchmark": "Tau2-bench Telecom", "category": "Agents", "rawScore": 98, "percentile": 62.5, "peerCount": 8, "sourceUrl": "https://openai.com/index/introducing-gpt-5-5/", "measuredAt": "2026-04-23" }, { "axisId": "toolathlon", "benchmark": "Toolathlon", "category": "Agents", "rawScore": 55.6, "percentile": 91.7, "peerCount": 6, "sourceUrl": "https://openai.com/index/introducing-gpt-5-5/", "measuredAt": "2026-04-23" }, { "axisId": "osworld-verified", "benchmark": "OSWorld-Verified", "category": "Agents", "rawScore": 78.7, "percentile": 62.5, "peerCount": 12, "sourceUrl": "https://openai.com/index/introducing-gpt-5-5/", "measuredAt": "2026-04-23" } ] }, { "modelId": "hy3", "name": "Hy3", "providerId": "tencent", "releaseDate": "2026-07-06", "rank": 8, "score": 66.8, "categoryScores": { "Coding": 66.6, "Reasoning": 68.8, "Agents": 65 }, "coverage": 7, "categoryCoverage": 3, "missingCategories": [], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 78, "percentile": 85.4, "peerCount": 48, "sourceUrl": "https://github.com/Tencent-Hunyuan/Hy3", "measuredAt": "2026-07-06" }, { "axisId": "swe-pro", "benchmark": "SWE-bench Pro", "category": "Coding", "rawScore": 57.9, "percentile": 47.7, "peerCount": 22, "sourceUrl": "https://github.com/Tencent-Hunyuan/Hy3", "measuredAt": "2026-07-06" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 90.4, "percentile": 70.3, "peerCount": 111, "sourceUrl": "https://github.com/Tencent-Hunyuan/Hy3", "measuredAt": "2026-07-06" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 37, "percentile": 67.2, "peerCount": 29, "sourceUrl": "https://github.com/Tencent-Hunyuan/Hy3", "measuredAt": "2026-07-06" }, { "axisId": "browsecomp", "benchmark": "BrowseComp", "category": "Agents", "rawScore": 84.2, "percentile": 70, "peerCount": 15, "sourceUrl": "https://github.com/Tencent-Hunyuan/Hy3", "measuredAt": "2026-07-06" }, { "axisId": "mcp-atlas", "benchmark": "MCP Atlas", "category": "Agents", "rawScore": 79.1, "percentile": 83.3, "peerCount": 9, "sourceUrl": "https://github.com/Tencent-Hunyuan/Hy3", "measuredAt": "2026-07-06" }, { "axisId": "toolathlon", "benchmark": "Toolathlon", "category": "Agents", "rawScore": 48.5, "percentile": 41.7, "peerCount": 6, "sourceUrl": "https://github.com/Tencent-Hunyuan/Hy3", "measuredAt": "2026-07-06" } ] }, { "modelId": "step-3-5-flash", "name": "Step 3.5 Flash", "providerId": "stepfun", "releaseDate": "2026-02-12", "rank": 9, "score": 59.8, "categoryScores": { "Coding": 64.9, "Reasoning": 91.3, "Agents": 23.3 }, "coverage": 4, "categoryCoverage": 3, "missingCategories": [], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 74.4, "percentile": 61.5, "peerCount": 48, "sourceUrl": "https://static.stepfun.com/blog/step-3.5-flash/", "measuredAt": "2026-02-12" }, { "axisId": "livecodebench-6", "benchmark": "LiveCodeBench v6", "category": "Coding", "rawScore": 86.4, "percentile": 68.2, "peerCount": 11, "sourceUrl": "https://static.stepfun.com/blog/step-3.5-flash/", "measuredAt": "2026-02-12" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 97.3, "percentile": 91.3, "peerCount": 40, "sourceUrl": "https://static.stepfun.com/blog/step-3.5-flash/", "measuredAt": "2026-02-12" }, { "axisId": "browsecomp", "benchmark": "BrowseComp", "category": "Agents", "rawScore": 51.6, "percentile": 23.3, "peerCount": 15, "sourceUrl": "https://static.stepfun.com/blog/step-3.5-flash/", "measuredAt": "2026-02-12" } ] }, { "modelId": "claude-sonnet-4-6", "name": "Claude Sonnet 4.6", "providerId": "anthropic", "releaseDate": "2026-02-17", "rank": 10, "score": 59.4, "categoryScores": { "Coding": 88.5, "Reasoning": 48.7, "Agents": 41 }, "coverage": 7, "categoryCoverage": 3, "missingCategories": [], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 79.6, "percentile": 88.5, "peerCount": 48, "sourceUrl": "https://www-cdn.anthropic.com/78073f739564e986ff3e28522761a7a0b4484f84.pdf", "measuredAt": "2026-02-17" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 89.9, "percentile": 67.6, "peerCount": 111, "sourceUrl": "https://www-cdn.anthropic.com/78073f739564e986ff3e28522761a7a0b4484f84.pdf", "measuredAt": "2026-02-17" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 33.2, "percentile": 53.4, "peerCount": 29, "sourceUrl": "https://www-cdn.anthropic.com/78073f739564e986ff3e28522761a7a0b4484f84.pdf", "measuredAt": "2026-02-17" }, { "axisId": "hle-with-tools", "benchmark": "Humanity's Last Exam (with tools)", "category": "Reasoning", "rawScore": 49, "percentile": 25, "peerCount": 14, "sourceUrl": "https://www-cdn.anthropic.com/78073f739564e986ff3e28522761a7a0b4484f84.pdf", "measuredAt": "2026-02-17" }, { "axisId": "mcp-atlas", "benchmark": "MCP Atlas", "category": "Agents", "rawScore": 61.3, "percentile": 50, "peerCount": 9, "sourceUrl": "https://www-cdn.anthropic.com/78073f739564e986ff3e28522761a7a0b4484f84.pdf", "measuredAt": "2026-02-17" }, { "axisId": "tau2-telecom", "benchmark": "Tau2-bench Telecom", "category": "Agents", "rawScore": 97.9, "percentile": 43.8, "peerCount": 8, "sourceUrl": "https://www-cdn.anthropic.com/78073f739564e986ff3e28522761a7a0b4484f84.pdf", "measuredAt": "2026-02-17" }, { "axisId": "osworld-verified", "benchmark": "OSWorld-Verified", "category": "Agents", "rawScore": 72.5, "percentile": 29.2, "peerCount": 12, "sourceUrl": "https://www-cdn.anthropic.com/78073f739564e986ff3e28522761a7a0b4484f84.pdf", "measuredAt": "2026-02-17" } ] }, { "modelId": "gpt-5-4", "name": "GPT-5.4", "providerId": "openai", "releaseDate": "2026-03-05", "rank": 11, "score": 59.3, "categoryScores": { "Coding": 43.2, "Reasoning": 69.5, "Agents": 65.3 }, "coverage": 10, "categoryCoverage": 3, "missingCategories": [], "components": [ { "axisId": "swe-pro", "benchmark": "SWE-bench Pro", "category": "Coding", "rawScore": 57.7, "percentile": 43.2, "peerCount": 22, "sourceUrl": "https://openai.com/index/introducing-gpt-5-4/", "measuredAt": "2026-03-05" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 92.8, "percentile": 85.6, "peerCount": 111, "sourceUrl": "https://openai.com/index/introducing-gpt-5-4/", "measuredAt": "2026-03-05" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 39.8, "percentile": 70.7, "peerCount": 29, "sourceUrl": "https://openai.com/index/introducing-gpt-5-4/", "measuredAt": "2026-03-05" }, { "axisId": "hle-with-tools", "benchmark": "Humanity's Last Exam (with tools)", "category": "Reasoning", "rawScore": 52.1, "percentile": 39.3, "peerCount": 14, "sourceUrl": "https://openai.com/index/introducing-gpt-5-4/", "measuredAt": "2026-03-05" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 78.6, "percentile": 82.2, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" }, { "axisId": "browsecomp", "benchmark": "BrowseComp", "category": "Agents", "rawScore": 82.7, "percentile": 63.3, "peerCount": 15, "sourceUrl": "https://openai.com/index/introducing-gpt-5-4/", "measuredAt": "2026-03-05" }, { "axisId": "mcp-atlas", "benchmark": "MCP Atlas", "category": "Agents", "rawScore": 67.2, "percentile": 61.1, "peerCount": 9, "sourceUrl": "https://openai.com/index/introducing-gpt-5-4/", "measuredAt": "2026-03-05" }, { "axisId": "tau2-telecom", "benchmark": "Tau2-bench Telecom", "category": "Agents", "rawScore": 98.9, "percentile": 81.3, "peerCount": 8, "sourceUrl": "https://openai.com/index/introducing-gpt-5-4/", "measuredAt": "2026-03-05" }, { "axisId": "toolathlon", "benchmark": "Toolathlon", "category": "Agents", "rawScore": 54.6, "percentile": 75, "peerCount": 6, "sourceUrl": "https://openai.com/index/introducing-gpt-5-4/", "measuredAt": "2026-03-05" }, { "axisId": "osworld-verified", "benchmark": "OSWorld-Verified", "category": "Agents", "rawScore": 75, "percentile": 45.8, "peerCount": 12, "sourceUrl": "https://openai.com/index/introducing-gpt-5-4/", "measuredAt": "2026-03-05" } ] }, { "modelId": "longcat-2-0", "name": "LongCat 2.0", "providerId": "meituan", "releaseDate": "2026-06-30", "rank": 11, "score": 59.3, "categoryScores": { "Coding": 56.8, "Reasoning": 64.4, "Agents": 56.7 }, "coverage": 3, "categoryCoverage": 3, "missingCategories": [], "components": [ { "axisId": "swe-pro", "benchmark": "SWE-bench Pro", "category": "Coding", "rawScore": 59.5, "percentile": 56.8, "peerCount": 22, "sourceUrl": "https://github.com/meituan-longcat/LongCat-2.0", "measuredAt": "2026-06-30" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 88.9, "percentile": 64.4, "peerCount": 111, "sourceUrl": "https://github.com/meituan-longcat/LongCat-2.0", "measuredAt": "2026-06-30" }, { "axisId": "browsecomp", "benchmark": "BrowseComp", "category": "Agents", "rawScore": 79.9, "percentile": 56.7, "peerCount": 15, "sourceUrl": "https://github.com/meituan-longcat/LongCat-2.0", "measuredAt": "2026-06-30" } ] }, { "modelId": "gemini-3-6-flash", "name": "Gemini 3.6 Flash", "providerId": "google", "releaseDate": "2026-07-21", "rank": 13, "score": 54.5, "categoryScores": { "Coding": 9.4, "Reasoning": 75, "Agents": 79.2 }, "coverage": 4, "categoryCoverage": 3, "missingCategories": [], "components": [ { "axisId": "deepswe-1-1", "benchmark": "DeepSWE v1.1", "category": "Coding", "rawScore": 49, "percentile": 9.4, "peerCount": 16, "sourceUrl": "https://deepmind.google/models/model-cards/gemini-3-6-flash/", "measuredAt": "2026-07-21" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 94.1, "percentile": 93.2, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 58.9, "percentile": 56.8, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" }, { "axisId": "osworld-verified", "benchmark": "OSWorld-Verified", "category": "Agents", "rawScore": 83, "percentile": 79.2, "peerCount": 12, "sourceUrl": "https://deepmind.google/models/model-cards/gemini-3-6-flash/", "measuredAt": "2026-07-21" } ] }, { "modelId": "dola-seed-2-0-pro", "name": "Dola Seed 2.0 Pro", "providerId": "bytedance", "releaseDate": "2026-02-16", "rank": 14, "score": 53.9, "categoryScores": { "Coding": 42.2, "Reasoning": 82.8, "Agents": 36.7 }, "coverage": 7, "categoryCoverage": 3, "missingCategories": [], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 76.5, "percentile": 72.9, "peerCount": 48, "sourceUrl": "https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf", "measuredAt": "2026-02-16" }, { "axisId": "swe-pro", "benchmark": "SWE-bench Pro", "category": "Coding", "rawScore": 46.9, "percentile": 11.4, "peerCount": 22, "sourceUrl": "https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf", "measuredAt": "2026-02-16" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 92.4, "percentile": 82, "peerCount": 111, "sourceUrl": "https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf", "measuredAt": "2026-02-16" }, { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 90.1, "percentile": 98.3, "peerCount": 29, "sourceUrl": "https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf", "measuredAt": "2026-02-16" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 99, "percentile": 93.8, "peerCount": 40, "sourceUrl": "https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf", "measuredAt": "2026-02-16" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 33.3, "percentile": 56.9, "peerCount": 29, "sourceUrl": "https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf", "measuredAt": "2026-02-16" }, { "axisId": "browsecomp", "benchmark": "BrowseComp", "category": "Agents", "rawScore": 77.3, "percentile": 36.7, "peerCount": 15, "sourceUrl": "https://lf3-static.bytednsdoc.com/obj/eden-cn/lapzild-tss/ljhwZthlaukjlkulzlp/seed2/0214/Seed2.0%20Model%20Card.pdf", "measuredAt": "2026-02-16" } ] }, { "modelId": "solar-open-2-250b", "name": "Solar Open 2 250B-A15B", "providerId": "upstage", "releaseDate": "2026-07-24", "rank": 15, "score": 53.5, "categoryScores": { "Coding": 69.1, "Reasoning": 63.5, "Agents": 27.8 }, "coverage": 6, "categoryCoverage": 3, "missingCategories": [], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 70.4, "percentile": 42.7, "peerCount": 48, "sourceUrl": "https://huggingface.co/upstage/Solar-Open2-250B", "measuredAt": "2026-07-24" }, { "axisId": "livecodebench-6", "benchmark": "LiveCodeBench v6", "category": "Coding", "rawScore": 92.4, "percentile": 95.5, "peerCount": 11, "sourceUrl": "https://huggingface.co/upstage/Solar-Open2-250B", "measuredAt": "2026-07-24" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 86.3, "percentile": 50.9, "peerCount": 111, "sourceUrl": "https://huggingface.co/upstage/Solar-Open2-250B", "measuredAt": "2026-07-24" }, { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 86.2, "percentile": 89.7, "peerCount": 29, "sourceUrl": "https://huggingface.co/upstage/Solar-Open2-250B", "measuredAt": "2026-07-24" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 28.8, "percentile": 50, "peerCount": 29, "sourceUrl": "https://huggingface.co/upstage/Solar-Open2-250B", "measuredAt": "2026-07-24" }, { "axisId": "mcp-atlas", "benchmark": "MCP Atlas", "category": "Agents", "rawScore": 58.2, "percentile": 27.8, "peerCount": 9, "sourceUrl": "https://huggingface.co/upstage/Solar-Open2-250B", "measuredAt": "2026-07-24" } ] }, { "modelId": "claude-sonnet-4-5", "name": "Claude Sonnet 4.5", "providerId": "anthropic", "releaseDate": "2025-09-29", "rank": 16, "score": 45.4, "categoryScores": { "Coding": 77.1, "Reasoning": 28.5, "Agents": 30.6 }, "coverage": 9, "categoryCoverage": 3, "missingCategories": [], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 77.2, "percentile": 77.1, "peerCount": 48, "sourceUrl": "https://www-cdn.anthropic.com/78073f739564e986ff3e28522761a7a0b4484f84.pdf", "measuredAt": "2026-02-17" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 83.4, "percentile": 42.8, "peerCount": 111, "sourceUrl": "https://www-cdn.anthropic.com/78073f739564e986ff3e28522761a7a0b4484f84.pdf", "measuredAt": "2026-02-17" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 84.2, "percentile": 36.3, "peerCount": 40, "sourceUrl": "https://matharena.ai/?comp=aime--aime_2025", "measuredAt": "2026-09-17" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 17.7, "percentile": 25.9, "peerCount": 29, "sourceUrl": "https://www-cdn.anthropic.com/78073f739564e986ff3e28522761a7a0b4484f84.pdf", "measuredAt": "2026-02-17" }, { "axisId": "hle-with-tools", "benchmark": "Humanity's Last Exam (with tools)", "category": "Reasoning", "rawScore": 33.6, "percentile": 17.9, "peerCount": 14, "sourceUrl": "https://www-cdn.anthropic.com/78073f739564e986ff3e28522761a7a0b4484f84.pdf", "measuredAt": "2026-02-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 23.9, "percentile": 19.5, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" }, { "axisId": "mcp-atlas", "benchmark": "MCP Atlas", "category": "Agents", "rawScore": 43.8, "percentile": 16.7, "peerCount": 9, "sourceUrl": "https://www-cdn.anthropic.com/78073f739564e986ff3e28522761a7a0b4484f84.pdf", "measuredAt": "2026-02-17" }, { "axisId": "tau2-telecom", "benchmark": "Tau2-bench Telecom", "category": "Agents", "rawScore": 98, "percentile": 62.5, "peerCount": 8, "sourceUrl": "https://www-cdn.anthropic.com/78073f739564e986ff3e28522761a7a0b4484f84.pdf", "measuredAt": "2026-02-17" }, { "axisId": "osworld-verified", "benchmark": "OSWorld-Verified", "category": "Agents", "rawScore": 61.4, "percentile": 12.5, "peerCount": 12, "sourceUrl": "https://www-cdn.anthropic.com/78073f739564e986ff3e28522761a7a0b4484f84.pdf", "measuredAt": "2026-02-17" } ] }, { "modelId": "gpt-5-4-mini", "name": "GPT-5.4 mini", "providerId": "openai", "releaseDate": "2026-03-17", "rank": 17, "score": 34.8, "categoryScores": { "Coding": 29.5, "Reasoning": 52, "Agents": 22.9 }, "coverage": 5, "categoryCoverage": 3, "missingCategories": [], "components": [ { "axisId": "swe-pro", "benchmark": "SWE-bench Pro", "category": "Coding", "rawScore": 54.4, "percentile": 29.5, "peerCount": 22, "sourceUrl": "https://openai.com/index/introducing-gpt-5-4-mini-and-nano/", "measuredAt": "2026-03-17" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 88, "percentile": 59.9, "peerCount": 111, "sourceUrl": "https://openai.com/index/introducing-gpt-5-4-mini-and-nano/", "measuredAt": "2026-03-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 51.2, "percentile": 44.1, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" }, { "axisId": "toolathlon", "benchmark": "Toolathlon", "category": "Agents", "rawScore": 42.9, "percentile": 25, "peerCount": 6, "sourceUrl": "https://openai.com/index/introducing-gpt-5-4-mini-and-nano/", "measuredAt": "2026-03-17" }, { "axisId": "osworld-verified", "benchmark": "OSWorld-Verified", "category": "Agents", "rawScore": 72.1, "percentile": 20.8, "peerCount": 12, "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", "releaseDate": "2026-06-04", "rank": 18, "score": 33.6, "categoryScores": { "Coding": 38.5, "Reasoning": 52.4, "Agents": 10 }, "coverage": 4, "categoryCoverage": 3, "missingCategories": [], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 69.7, "percentile": 38.5, "peerCount": 48, "sourceUrl": "https://build.nvidia.com/nvidia/nemotron-3-ultra-550b-a55b/modelcard", "measuredAt": "2026-06-04" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 87.9, "percentile": 58.1, "peerCount": 111, "sourceUrl": "https://build.nvidia.com/nvidia/nemotron-3-ultra-550b-a55b/modelcard", "measuredAt": "2026-06-04" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 26.1, "percentile": 46.6, "peerCount": 29, "sourceUrl": "https://build.nvidia.com/nvidia/nemotron-3-ultra-550b-a55b/modelcard", "measuredAt": "2026-06-04" }, { "axisId": "browsecomp", "benchmark": "BrowseComp", "category": "Agents", "rawScore": 41.4, "percentile": 10, "peerCount": 15, "sourceUrl": "https://build.nvidia.com/nvidia/nemotron-3-ultra-550b-a55b/modelcard", "measuredAt": "2026-06-04" } ] }, { "modelId": "gpt-5-4-nano", "name": "GPT-5.4 nano", "providerId": "openai", "releaseDate": "2026-03-17", "rank": 19, "score": 20.1, "categoryScores": { "Coding": 15.9, "Reasoning": 38, "Agents": 6.3 }, "coverage": 5, "categoryCoverage": 3, "missingCategories": [], "components": [ { "axisId": "swe-pro", "benchmark": "SWE-bench Pro", "category": "Coding", "rawScore": 52.4, "percentile": 15.9, "peerCount": 22, "sourceUrl": "https://openai.com/index/introducing-gpt-5-4-mini-and-nano/", "measuredAt": "2026-03-17" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 82.8, "percentile": 38.7, "peerCount": 111, "sourceUrl": "https://openai.com/index/introducing-gpt-5-4-mini-and-nano/", "measuredAt": "2026-03-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 44.9, "percentile": 37.3, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" }, { "axisId": "toolathlon", "benchmark": "Toolathlon", "category": "Agents", "rawScore": 35.5, "percentile": 8.3, "peerCount": 6, "sourceUrl": "https://openai.com/index/introducing-gpt-5-4-mini-and-nano/", "measuredAt": "2026-03-17" }, { "axisId": "osworld-verified", "benchmark": "OSWorld-Verified", "category": "Agents", "rawScore": 39, "percentile": 4.2, "peerCount": 12, "sourceUrl": "https://openai.com/index/introducing-gpt-5-4-mini-and-nano/", "measuredAt": "2026-03-17" } ] }, { "modelId": "nvidia-nemotron-3-5-lightning-30b-a3b-nvfp4", "name": "NVIDIA Nemotron 3.5 Lightning 30B-A3B NVFP4", "providerId": "nvidia", "releaseDate": "2026-08-11", "rank": 20, "score": 15.4, "categoryScores": { "Coding": 13.5, "Reasoning": 29.3, "Agents": 3.3 }, "coverage": 5, "categoryCoverage": 3, "missingCategories": [], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 52.8, "percentile": 13.5, "peerCount": 48, "sourceUrl": "https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-NVFP4", "measuredAt": "2026-08-11" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 75.57, "percentile": 25.7, "peerCount": 111, "sourceUrl": "https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-NVFP4", "measuredAt": "2026-08-11" }, { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 81.62, "percentile": 50, "peerCount": 29, "sourceUrl": "https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-NVFP4", "measuredAt": "2026-08-11" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 10.47, "percentile": 12.1, "peerCount": 29, "sourceUrl": "https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-NVFP4", "measuredAt": "2026-08-11" }, { "axisId": "browsecomp", "benchmark": "BrowseComp", "category": "Agents", "rawScore": 36.81, "percentile": 3.3, "peerCount": 15, "sourceUrl": "https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-NVFP4", "measuredAt": "2026-08-11" } ] } ], "unranked": [ { "modelId": "gpt-6-sol", "name": "GPT-6 Sol", "providerId": "openai", "releaseDate": "2026-09-22", "rank": null, "score": null, "categoryScores": { "Coding": 65.6, "Reasoning": null, "Agents": 16.7 }, "coverage": 2, "categoryCoverage": 2, "missingCategories": [ "Reasoning" ], "components": [ { "axisId": "deepswe-1-1", "benchmark": "DeepSWE v1.1", "category": "Coding", "rawScore": 68.8, "percentile": 65.6, "peerCount": 16, "sourceUrl": "https://openai.com/index/introducing-gpt-6-sol-and-luna/", "measuredAt": "2026-09-22" }, { "axisId": "automation-1-0-6", "benchmark": "AutomationBench 1.0.6", "category": "Agents", "rawScore": 33.2, "percentile": 16.7, "peerCount": 3, "sourceUrl": "https://zapier.com/benchmarks", "measuredAt": "2026-09-22" } ] }, { "modelId": "gpt-6-luna", "name": "GPT-6 Luna", "providerId": "openai", "releaseDate": "2026-09-22", "rank": null, "score": null, "categoryScores": { "Coding": 40.6, "Reasoning": null, "Agents": null }, "coverage": 1, "categoryCoverage": 1, "missingCategories": [ "Reasoning", "Agents" ], "components": [ { "axisId": "deepswe-1-1", "benchmark": "DeepSWE v1.1", "category": "Coding", "rawScore": 66.6, "percentile": 40.6, "peerCount": 16, "sourceUrl": "https://openai.com/index/introducing-gpt-6-sol-and-luna/", "measuredAt": "2026-09-22" } ] }, { "modelId": "grok-4-7", "name": "Grok 4.7", "providerId": "xai", "releaseDate": "2026-09-21", "rank": null, "score": null, "categoryScores": { "Coding": 64.1, "Reasoning": null, "Agents": null }, "coverage": 2, "categoryCoverage": 1, "missingCategories": [ "Reasoning", "Agents" ], "components": [ { "axisId": "deepswe-1-1", "benchmark": "DeepSWE v1.1", "category": "Coding", "rawScore": 71, "percentile": 78.1, "peerCount": 16, "sourceUrl": "https://x.ai/news/grok-4-7", "measuredAt": "2026-09-21" }, { "axisId": "terminal-4", "benchmark": "Terminal-Bench 4.0", "category": "Coding", "rawScore": 38, "percentile": 50, "peerCount": 3, "sourceUrl": "https://x.ai/news/grok-4-7", "measuredAt": "2026-09-21" } ] }, { "modelId": "glm-5-3-flash", "name": "GLM-5.3-Flash", "providerId": "zai", "releaseDate": "2026-09-16", "rank": null, "score": null, "categoryScores": { "Coding": 21.9, "Reasoning": 59.9, "Agents": null }, "coverage": 3, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "deepswe-1-1", "benchmark": "DeepSWE v1.1", "category": "Coding", "rawScore": 63.4, "percentile": 21.9, "peerCount": 16, "sourceUrl": "https://autoclaw.z.ai/blog/model/glm-5.3-flash/", "measuredAt": "2026-09-16" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 90.2, "percentile": 68.9, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 55.8, "percentile": 50.8, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "deepseek-v4-1-flash", "name": "DeepSeek V4.1 Flash", "providerId": "deepseek", "releaseDate": "2026-09-10", "rank": null, "score": null, "categoryScores": { "Coding": 96.9, "Reasoning": 75.7, "Agents": null }, "coverage": 2, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "deepswe-1-1", "benchmark": "DeepSWE v1.1", "category": "Coding", "rawScore": 74.2, "percentile": 96.9, "peerCount": 16, "sourceUrl": "https://api-docs.deepseek.com/updates/", "measuredAt": "2026-09-10" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 90.9, "percentile": 75.7, "peerCount": 111, "sourceUrl": "https://api-docs.deepseek.com/updates/", "measuredAt": "2026-09-10" } ] }, { "modelId": "gemini-3-8-flash", "name": "Gemini 3.8 Flash", "providerId": "google", "releaseDate": "2026-09-02", "rank": null, "score": null, "categoryScores": { "Coding": 53.7, "Reasoning": 84.4, "Agents": null }, "coverage": 4, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "deepswe-1-1", "benchmark": "DeepSWE v1.1", "category": "Coding", "rawScore": 73.7, "percentile": 90.6, "peerCount": 16, "sourceUrl": "https://deepmind.google/models/model-cards/gemini-3-8-flash/", "measuredAt": "2026-09-02" }, { "axisId": "terminal-4", "benchmark": "Terminal-Bench 4.0", "category": "Coding", "rawScore": 19.1, "percentile": 16.7, "peerCount": 3, "sourceUrl": "https://deepmind.google/models/model-cards/gemini-3-8-flash/", "measuredAt": "2026-09-02" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 95.3, "percentile": 98.6, "peerCount": 111, "sourceUrl": "https://openai.com/index/gpt-6-astra/", "measuredAt": "2026-09-03" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 68.4, "percentile": 70.3, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "claude-fable-5-1", "name": "Claude Fable 5.1", "providerId": "anthropic", "releaseDate": "2026-09-01", "rank": null, "score": null, "categoryScores": { "Coding": 37.5, "Reasoning": 97.9, "Agents": null }, "coverage": 3, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "terminal-science-0-1", "benchmark": "Terminal-Bench Science 0.1", "category": "Coding", "rawScore": 52.6, "percentile": 37.5, "peerCount": 4, "sourceUrl": "https://www.anthropic.com/claude-fable-and-mythos-5-1", "measuredAt": "2026-09-01" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 60.9, "percentile": 98.3, "peerCount": 29, "sourceUrl": "https://www.anthropic.com/claude-fable-and-mythos-5-1", "measuredAt": "2026-09-01" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 90.2, "percentile": 97.5, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "qwen3-8-flash", "name": "Qwen3.8-Flash", "providerId": "qwen", "releaseDate": "2026-08-26", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "solar-pro-4", "name": "Solar Pro 4", "providerId": "upstage", "releaseDate": "2026-08-20", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "glm-5-3", "name": "GLM-5.3", "providerId": "zai", "releaseDate": "2026-08-14", "rank": null, "score": null, "categoryScores": { "Coding": 46.9, "Reasoning": 73.9, "Agents": null }, "coverage": 3, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "deepswe-1-1", "benchmark": "DeepSWE v1.1", "category": "Coding", "rawScore": 66.9, "percentile": 46.9, "peerCount": 16, "sourceUrl": "https://huggingface.co/zai-org/GLM-5.3", "measuredAt": "2026-08-14" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 90.9, "percentile": 75.7, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 68.8, "percentile": 72, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "qwen3-8-27b", "name": "Qwen3.8 27B", "providerId": "qwen", "releaseDate": "2026-08-14", "rank": null, "score": null, "categoryScores": { "Coding": 61.4, "Reasoning": 65.3, "Agents": null }, "coverage": 2, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-pro", "benchmark": "SWE-bench Pro", "category": "Coding", "rawScore": 61.7, "percentile": 61.4, "peerCount": 22, "sourceUrl": "https://huggingface.co/Qwen/Qwen3.8-27B", "measuredAt": "2026-08-14" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 89.2, "percentile": 65.3, "peerCount": 111, "sourceUrl": "https://huggingface.co/Qwen/Qwen3.8-27B", "measuredAt": "2026-08-14" } ] }, { "modelId": "deepseek-v4-pro", "name": "DeepSeek V4 Pro", "providerId": "deepseek", "releaseDate": "2026-08-13", "rank": null, "score": null, "categoryScores": { "Coding": 15.6, "Reasoning": 68.3, "Agents": null }, "coverage": 4, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "deepswe-1-1", "benchmark": "DeepSWE v1.1", "category": "Coding", "rawScore": 62.7, "percentile": 15.6, "peerCount": 16, "sourceUrl": "https://api-docs.deepseek.com/updates/", "measuredAt": "2026-08-13" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 90.9, "percentile": 75.7, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "hle-with-tools", "benchmark": "Humanity's Last Exam (with tools)", "category": "Reasoning", "rawScore": 60, "percentile": 89.3, "peerCount": 14, "sourceUrl": "https://api-docs.deepseek.com/updates/", "measuredAt": "2026-08-13" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 45.3, "percentile": 39.8, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "gemini-3-7-flash", "name": "Gemini 3.7 Flash", "providerId": "google", "releaseDate": "2026-08-13", "rank": null, "score": null, "categoryScores": { "Coding": 28.1, "Reasoning": 86.6, "Agents": null }, "coverage": 3, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "deepswe-1-1", "benchmark": "DeepSWE v1.1", "category": "Coding", "rawScore": 65.3, "percentile": 28.1, "peerCount": 16, "sourceUrl": "https://deepmind.google/models/model-cards/gemini-3-7-flash/", "measuredAt": "2026-08-13" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 94.8, "percentile": 97.7, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 71.6, "percentile": 75.4, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "mai-thinking-1", "name": "MAI-Thinking-1", "providerId": "microsoft", "releaseDate": "2026-08-12", "rank": null, "score": null, "categoryScores": { "Coding": 52.4, "Reasoning": 68.6, "Agents": null }, "coverage": 6, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 73.5, "percentile": 59.4, "peerCount": 48, "sourceUrl": "https://microsoft.ai/pdf/mai-thinking-1.pdf", "measuredAt": "2026-08-12" }, { "axisId": "swe-pro", "benchmark": "SWE-bench Pro", "category": "Coding", "rawScore": 52.8, "percentile": 20.5, "peerCount": 22, "sourceUrl": "https://microsoft.ai/pdf/mai-thinking-1.pdf", "measuredAt": "2026-08-12" }, { "axisId": "livecodebench-6", "benchmark": "LiveCodeBench v6", "category": "Coding", "rawScore": 87.7, "percentile": 77.3, "peerCount": 11, "sourceUrl": "https://microsoft.ai/pdf/mai-thinking-1.pdf", "measuredAt": "2026-08-12" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 84.2, "percentile": 44.6, "peerCount": 111, "sourceUrl": "https://microsoft.ai/pdf/mai-thinking-1.pdf", "measuredAt": "2026-08-12" }, { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 85, "percentile": 72.4, "peerCount": 29, "sourceUrl": "https://microsoft.ai/pdf/mai-thinking-1.pdf", "measuredAt": "2026-08-12" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 97, "percentile": 88.8, "peerCount": 40, "sourceUrl": "https://microsoft.ai/pdf/mai-thinking-1.pdf", "measuredAt": "2026-08-12" } ] }, { "modelId": "grok-4-6", "name": "Grok 4.6", "providerId": "xai", "releaseDate": "2026-08-12", "rank": null, "score": null, "categoryScores": { "Coding": 34.4, "Reasoning": 79.2, "Agents": null }, "coverage": 3, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "deepswe-1-1", "benchmark": "DeepSWE v1.1", "category": "Coding", "rawScore": 65.9, "percentile": 34.4, "peerCount": 16, "sourceUrl": "https://x.ai/news/grok-4-6", "measuredAt": "2026-08-12" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 94, "percentile": 92.3, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 66, "percentile": 66.1, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "qwen3-8-2-4t-a95b", "name": "Qwen3.8 2.4T-A95B", "providerId": "qwen", "releaseDate": "2026-08-12", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 83.8, "Agents": null }, "coverage": 1, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 92.6, "percentile": 83.8, "peerCount": 111, "sourceUrl": "https://docs.modelstudio.console.alibabacloud.com/en/model-studio/qwen3-8-2-4t-a95b", "measuredAt": "2026-08-12" } ] }, { "modelId": "qwen3-8-max", "name": "Qwen3.8-Max", "providerId": "qwen", "releaseDate": "2026-08-02", "rank": null, "score": null, "categoryScores": { "Coding": 93.2, "Reasoning": 82.2, "Agents": null }, "coverage": 3, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-pro", "benchmark": "SWE-bench Pro", "category": "Coding", "rawScore": 67.7, "percentile": 93.2, "peerCount": 22, "sourceUrl": "https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B", "measuredAt": "2026-08-12" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 92.6, "percentile": 83.8, "peerCount": 111, "sourceUrl": "https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B", "measuredAt": "2026-08-12" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 74.7, "percentile": 80.5, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "claude-opus-5", "name": "Claude Opus 5", "providerId": "anthropic", "releaseDate": "2026-07-24", "rank": null, "score": null, "categoryScores": { "Coding": 12.5, "Reasoning": 92.3, "Agents": null }, "coverage": 4, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "terminal-science-0-1", "benchmark": "Terminal-Bench Science 0.1", "category": "Coding", "rawScore": 29, "percentile": 12.5, "peerCount": 4, "sourceUrl": "https://www.anthropic.com/claude-fable-and-mythos-5-1", "measuredAt": "2026-09-01" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 93.9, "percentile": 91.4, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 56.6, "percentile": 94.8, "peerCount": 29, "sourceUrl": "https://www.anthropic.com/claude-fable-and-mythos-5-1", "measuredAt": "2026-09-01" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 85.6, "percentile": 90.7, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "gemini-3-5-flash-lite", "name": "Gemini 3.5 Flash-Lite", "providerId": "google", "releaseDate": "2026-07-21", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 32, "Agents": null }, "coverage": 2, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 83.3, "percentile": 41, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 26, "percentile": 22.9, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "qwen3-7-flash", "name": "Qwen3.7-Flash", "providerId": "qwen", "releaseDate": "2026-07-15", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 25.9, "Agents": null }, "coverage": 2, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 82.3, "percentile": 35.6, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 19.3, "percentile": 16.1, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "dola-seed-2-1-turbo", "name": "Dola Seed 2.1 Turbo", "providerId": "bytedance", "releaseDate": "2026-07-13", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "seed-2-1-pro", "name": "Seed2.1 Pro", "providerId": "bytedance", "releaseDate": "2026-07-13", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "gpt-5-6-luna", "name": "GPT-5.6 Luna", "providerId": "openai", "releaseDate": "2026-07-09", "rank": null, "score": null, "categoryScores": { "Coding": 64.1, "Reasoning": 83.1, "Agents": null }, "coverage": 4, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-pro", "benchmark": "SWE-bench Pro", "category": "Coding", "rawScore": 62.7, "percentile": 75, "peerCount": 22, "sourceUrl": "https://openai.com/index/gpt-5-6/", "measuredAt": "2026-07-09" }, { "axisId": "deepswe-1-1", "benchmark": "DeepSWE v1.1", "category": "Coding", "rawScore": 67.2, "percentile": 53.1, "peerCount": 16, "sourceUrl": "https://openai.com/index/gpt-5-6/", "measuredAt": "2026-07-09" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 92.3, "percentile": 80.6, "peerCount": 111, "sourceUrl": "https://openai.com/index/gpt-5-6/", "measuredAt": "2026-07-09" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 82.1, "percentile": 85.6, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "gpt-5-6-sol", "name": "GPT-5.6 Sol", "providerId": "openai", "releaseDate": "2026-07-09", "rank": null, "score": null, "categoryScores": { "Coding": 86.5, "Reasoning": 96.3, "Agents": null }, "coverage": 4, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-pro", "benchmark": "SWE-bench Pro", "category": "Coding", "rawScore": 64.6, "percentile": 88.6, "peerCount": 22, "sourceUrl": "https://openai.com/index/gpt-5-6/", "measuredAt": "2026-07-09" }, { "axisId": "deepswe-1-1", "benchmark": "DeepSWE v1.1", "category": "Coding", "rawScore": 72.7, "percentile": 84.4, "peerCount": 16, "sourceUrl": "https://openai.com/index/gpt-5-6/", "measuredAt": "2026-07-09" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 94.6, "percentile": 96.8, "peerCount": 111, "sourceUrl": "https://openai.com/index/gpt-5-6/", "measuredAt": "2026-07-09" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 89.1, "percentile": 95.8, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "gpt-5-6-terra", "name": "GPT-5.6 Terra", "providerId": "openai", "releaseDate": "2026-07-09", "rank": null, "score": null, "categoryScores": { "Coding": 75.7, "Reasoning": 89.7, "Agents": null }, "coverage": 4, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-pro", "benchmark": "SWE-bench Pro", "category": "Coding", "rawScore": 63.4, "percentile": 79.5, "peerCount": 22, "sourceUrl": "https://openai.com/index/gpt-5-6/", "measuredAt": "2026-07-09" }, { "axisId": "deepswe-1-1", "benchmark": "DeepSWE v1.1", "category": "Coding", "rawScore": 69.6, "percentile": 71.9, "peerCount": 16, "sourceUrl": "https://openai.com/index/gpt-5-6/", "measuredAt": "2026-07-09" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 92.9, "percentile": 86.9, "peerCount": 111, "sourceUrl": "https://openai.com/index/gpt-5-6/", "measuredAt": "2026-07-09" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 86, "percentile": 92.4, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "grok-4-5", "name": "Grok 4.5", "providerId": "xai", "releaseDate": "2026-07-08", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 71.5, "Agents": null }, "coverage": 2, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 93.4, "percentile": 88.7, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 57.2, "percentile": 54.2, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "claude-sonnet-5", "name": "Claude Sonnet 5", "providerId": "anthropic", "releaseDate": "2026-06-30", "rank": null, "score": null, "categoryScores": { "Coding": 94.8, "Reasoning": 71, "Agents": null }, "coverage": 5, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 85.2, "percentile": 94.8, "peerCount": 48, "sourceUrl": "https://www-cdn.anthropic.com/283ef97c476cf442c91d9a37d5b214242a55bb92/Claude%20Sonnet%205%20System%20Card.pdf", "measuredAt": "2026-06-30" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 90.5, "percentile": 71.6, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 43.2, "percentile": 81, "peerCount": 29, "sourceUrl": "https://www-cdn.anthropic.com/283ef97c476cf442c91d9a37d5b214242a55bb92/Claude%20Sonnet%205%20System%20Card.pdf", "measuredAt": "2026-06-30" }, { "axisId": "hle-with-tools", "benchmark": "Humanity's Last Exam (with tools)", "category": "Reasoning", "rawScore": 57.4, "percentile": 67.9, "peerCount": 14, "sourceUrl": "https://www-cdn.anthropic.com/283ef97c476cf442c91d9a37d5b214242a55bb92/Claude%20Sonnet%205%20System%20Card.pdf", "measuredAt": "2026-06-30" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 65.6, "percentile": 63.6, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "glm-5-2", "name": "GLM-5.2", "providerId": "zai", "releaseDate": "2026-06-24", "rank": null, "score": null, "categoryScores": { "Coding": 34.5, "Reasoning": 69.1, "Agents": null }, "coverage": 4, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-pro", "benchmark": "SWE-bench Pro", "category": "Coding", "rawScore": 62.1, "percentile": 65.9, "peerCount": 22, "sourceUrl": "https://docs.z.ai/guides/llm/glm-5.2", "measuredAt": "2026-06-24" }, { "axisId": "deepswe-1-1", "benchmark": "DeepSWE v1.1", "category": "Coding", "rawScore": 46.2, "percentile": 3.1, "peerCount": 16, "sourceUrl": "https://huggingface.co/zai-org/GLM-5.3", "measuredAt": "2026-08-14" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 91.9, "percentile": 79.7, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 59.2, "percentile": 58.5, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "kimi-k2-7-code", "name": "Kimi K2.7 Code", "providerId": "moonshot", "releaseDate": "2026-06-12", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 52.4, "Agents": null }, "coverage": 2, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 87.9, "percentile": 58.1, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 54, "percentile": 46.6, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "north-mini-code", "name": "North Mini Code", "providerId": "cohere", "releaseDate": "2026-06-09", "rank": null, "score": null, "categoryScores": { "Coding": 18.4, "Reasoning": null, "Agents": null }, "coverage": 3, "categoryCoverage": 1, "missingCategories": [ "Reasoning", "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 67.6, "percentile": 30.2, "peerCount": 48, "sourceUrl": "https://cohere.com/blog/north-mini-code", "measuredAt": "2026-06-09" }, { "axisId": "swe-pro", "benchmark": "SWE-bench Pro", "category": "Coding", "rawScore": 40.2, "percentile": 2.3, "peerCount": 22, "sourceUrl": "https://cohere.com/blog/north-mini-code", "measuredAt": "2026-06-09" }, { "axisId": "livecodebench-6", "benchmark": "LiveCodeBench v6", "category": "Coding", "rawScore": 70.3, "percentile": 22.7, "peerCount": 11, "sourceUrl": "https://cohere.com/blog/north-mini-code", "measuredAt": "2026-06-09" } ] }, { "modelId": "gemma-4-12b", "name": "Gemma 4 12B Unified", "providerId": "google", "releaseDate": "2026-06-03", "rank": null, "score": null, "categoryScores": { "Coding": 31.8, "Reasoning": 21, "Agents": null }, "coverage": 4, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "livecodebench-6", "benchmark": "LiveCodeBench v6", "category": "Coding", "rawScore": 72, "percentile": 31.8, "peerCount": 11, "sourceUrl": "https://ai.google.dev/gemma/docs/core/model_card_4", "measuredAt": "2026-06-03" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 78.8, "percentile": 28.4, "peerCount": 111, "sourceUrl": "https://ai.google.dev/gemma/docs/core/model_card_4", "measuredAt": "2026-06-03" }, { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 77.2, "percentile": 32.8, "peerCount": 29, "sourceUrl": "https://ai.google.dev/gemma/docs/core/model_card_4", "measuredAt": "2026-06-03" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 5.2, "percentile": 1.7, "peerCount": 29, "sourceUrl": "https://ai.google.dev/gemma/docs/core/model_card_4", "measuredAt": "2026-06-03" } ] }, { "modelId": "minimax-m3", "name": "MiniMax M3", "providerId": "minimax", "releaseDate": "2026-06-01", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 75.7, "Agents": null }, "coverage": 1, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 90.9, "percentile": 75.7, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" } ] }, { "modelId": "step-3-7-flash", "name": "Step 3.7 Flash", "providerId": "stepfun", "releaseDate": "2026-05-29", "rank": null, "score": null, "categoryScores": { "Coding": 55.8, "Reasoning": null, "Agents": 44.2 }, "coverage": 4, "categoryCoverage": 2, "missingCategories": [ "Reasoning" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 76.5, "percentile": 72.9, "peerCount": 48, "sourceUrl": "https://static.stepfun.com/blog/step-3.7-flash/", "measuredAt": "2026-05-29" }, { "axisId": "swe-pro", "benchmark": "SWE-bench Pro", "category": "Coding", "rawScore": 56.3, "percentile": 38.6, "peerCount": 22, "sourceUrl": "https://static.stepfun.com/blog/step-3.7-flash/", "measuredAt": "2026-05-29" }, { "axisId": "browsecomp", "benchmark": "BrowseComp", "category": "Agents", "rawScore": 75.8, "percentile": 30, "peerCount": 15, "sourceUrl": "https://static.stepfun.com/blog/step-3.7-flash/", "measuredAt": "2026-05-29" }, { "axisId": "toolathlon", "benchmark": "Toolathlon", "category": "Agents", "rawScore": 49.5, "percentile": 58.3, "peerCount": 6, "sourceUrl": "https://static.stepfun.com/blog/step-3.7-flash/", "measuredAt": "2026-05-29" } ] }, { "modelId": "qwen3-7-plus", "name": "Qwen3.7-Plus", "providerId": "qwen", "releaseDate": "2026-05-26", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 58.1, "Agents": null }, "coverage": 1, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 87.9, "percentile": 58.1, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" } ] }, { "modelId": "command-a-plus", "name": "Command A+", "providerId": "cohere", "releaseDate": "2026-05-20", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 60, "Agents": 31.3 }, "coverage": 2, "categoryCoverage": 2, "missingCategories": [ "Coding" ], "components": [ { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 90, "percentile": 60, "peerCount": 40, "sourceUrl": "https://cohere.com/blog/command-a-plus", "measuredAt": "2026-05-20" }, { "axisId": "tau2-telecom", "benchmark": "Tau2-bench Telecom", "category": "Agents", "rawScore": 85, "percentile": 31.3, "peerCount": 8, "sourceUrl": "https://cohere.com/blog/command-a-plus", "measuredAt": "2026-05-20" } ] }, { "modelId": "qwen3-7-max", "name": "Qwen3.7-Max", "providerId": "qwen", "releaseDate": "2026-05-20", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "gemini-3-5-flash", "name": "Gemini 3.5 Flash", "providerId": "google", "releaseDate": "2026-05-19", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 73.8, "Agents": 74.3 }, "coverage": 4, "categoryCoverage": 2, "missingCategories": [ "Coding" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 92.8, "percentile": 85.6, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 62.8, "percentile": 61.9, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" }, { "axisId": "mcp-atlas", "benchmark": "MCP Atlas", "category": "Agents", "rawScore": 83.6, "percentile": 94.4, "peerCount": 9, "sourceUrl": "https://deepmind.google/models/model-cards/gemini-3-5-flash/", "measuredAt": "2026-05-19" }, { "axisId": "osworld-verified", "benchmark": "OSWorld-Verified", "category": "Agents", "rawScore": 78.4, "percentile": 54.2, "peerCount": 12, "sourceUrl": "https://deepmind.google/models/model-cards/gemini-3-5-flash/", "measuredAt": "2026-05-19" } ] }, { "modelId": "grok-build-0-1", "name": "Grok Build 0.1", "providerId": "xai", "releaseDate": "2026-05-19", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "ernie-5-1", "name": "ERNIE 5.1", "providerId": "baidu", "releaseDate": "2026-05-09", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "mistral-medium-3-5", "name": "Mistral Medium 3.5", "providerId": "mistral", "releaseDate": "2026-04-28", "rank": null, "score": null, "categoryScores": { "Coding": 80.2, "Reasoning": null, "Agents": null }, "coverage": 1, "categoryCoverage": 1, "missingCategories": [ "Reasoning", "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 77.6, "percentile": 80.2, "peerCount": 48, "sourceUrl": "https://mistral.ai/news/vibe-remote-agents-mistral-medium-3-5/", "measuredAt": "2026-05-22" } ] }, { "modelId": "gpt-5-5-pro", "name": "GPT-5.5 Pro", "providerId": "openai", "releaseDate": "2026-04-23", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 77.5, "Agents": 96.7 }, "coverage": 4, "categoryCoverage": 2, "missingCategories": [ "Coding" ], "components": [ { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 43.1, "percentile": 77.6, "peerCount": 29, "sourceUrl": "https://openai.com/index/introducing-gpt-5-5/", "measuredAt": "2026-04-23" }, { "axisId": "hle-with-tools", "benchmark": "Humanity's Last Exam (with tools)", "category": "Reasoning", "rawScore": 57.2, "percentile": 60.7, "peerCount": 14, "sourceUrl": "https://openai.com/index/introducing-gpt-5-5/", "measuredAt": "2026-04-23" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 87.7, "percentile": 94.1, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" }, { "axisId": "browsecomp", "benchmark": "BrowseComp", "category": "Agents", "rawScore": 90.1, "percentile": 96.7, "peerCount": 15, "sourceUrl": "https://openai.com/index/introducing-gpt-5-5/", "measuredAt": "2026-04-23" } ] }, { "modelId": "qwen3-6-27b", "name": "Qwen3.6 27B", "providerId": "qwen", "releaseDate": "2026-04-21", "rank": null, "score": null, "categoryScores": { "Coding": 77.1, "Reasoning": 58.4, "Agents": null }, "coverage": 4, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 77.2, "percentile": 77.1, "peerCount": 48, "sourceUrl": "https://qwen.ai/blog?id=qwen3.6-27b", "measuredAt": "2026-04-21" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 87.8, "percentile": 55.9, "peerCount": 111, "sourceUrl": "https://qwen.ai/blog?id=qwen3.6-27b", "measuredAt": "2026-04-21" }, { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 86.2, "percentile": 89.7, "peerCount": 29, "sourceUrl": "https://qwen.ai/blog?id=qwen3.6-27b", "measuredAt": "2026-04-21" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 35.1, "percentile": 29.7, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "kimi-k2-6", "name": "Kimi K2.6", "providerId": "moonshot", "releaseDate": "2026-04-20", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 63.8, "Agents": null }, "coverage": 2, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 90.8, "percentile": 73.4, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 57.2, "percentile": 54.2, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "qwen3-6-flash", "name": "Qwen3.6-Flash", "providerId": "qwen", "releaseDate": "2026-04-16", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 26, "Agents": null }, "coverage": 2, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 83.3, "percentile": 41, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 17.2, "percentile": 11, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "glm-5-1", "name": "GLM-5.1", "providerId": "zai", "releaseDate": "2026-04-07", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 49.5, "Agents": null }, "coverage": 2, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 89.9, "percentile": 67.6, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 36.8, "percentile": 31.4, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "gemma-4-26b-a4b", "name": "Gemma 4 26B A4B", "providerId": "google", "releaseDate": "2026-04-02", "rank": null, "score": null, "categoryScores": { "Coding": 50, "Reasoning": 32.5, "Agents": null }, "coverage": 4, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "livecodebench-6", "benchmark": "LiveCodeBench v6", "category": "Coding", "rawScore": 77.1, "percentile": 50, "peerCount": 11, "sourceUrl": "https://ai.google.dev/gemma/docs/core/model_card_4", "measuredAt": "2026-04-02" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 82.3, "percentile": 35.6, "peerCount": 111, "sourceUrl": "https://ai.google.dev/gemma/docs/core/model_card_4", "measuredAt": "2026-04-02" }, { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 82.6, "percentile": 53.4, "peerCount": 29, "sourceUrl": "https://ai.google.dev/gemma/docs/core/model_card_4", "measuredAt": "2026-04-02" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 8.7, "percentile": 8.6, "peerCount": 29, "sourceUrl": "https://ai.google.dev/gemma/docs/core/model_card_4", "measuredAt": "2026-04-02" } ] }, { "modelId": "gemma-4-31b", "name": "Gemma 4 31B", "providerId": "google", "releaseDate": "2026-04-02", "rank": null, "score": null, "categoryScores": { "Coding": 59.1, "Reasoning": 52.5, "Agents": null }, "coverage": 4, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "livecodebench-6", "benchmark": "LiveCodeBench v6", "category": "Coding", "rawScore": 80, "percentile": 59.1, "peerCount": 11, "sourceUrl": "https://ai.google.dev/gemma/docs/core/model_card_4", "measuredAt": "2026-04-02" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 84.3, "percentile": 45.5, "peerCount": 111, "sourceUrl": "https://ai.google.dev/gemma/docs/core/model_card_4", "measuredAt": "2026-04-02" }, { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 85.2, "percentile": 79.3, "peerCount": 29, "sourceUrl": "https://ai.google.dev/gemma/docs/core/model_card_4", "measuredAt": "2026-04-02" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 19.5, "percentile": 32.8, "peerCount": 29, "sourceUrl": "https://ai.google.dev/gemma/docs/core/model_card_4", "measuredAt": "2026-04-02" } ] }, { "modelId": "gemma-4-e2b", "name": "Gemma 4 E2B", "providerId": "google", "releaseDate": "2026-04-02", "rank": null, "score": null, "categoryScores": { "Coding": 4.5, "Reasoning": 3.3, "Agents": null }, "coverage": 3, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "livecodebench-6", "benchmark": "LiveCodeBench v6", "category": "Coding", "rawScore": 44, "percentile": 4.5, "peerCount": 11, "sourceUrl": "https://ai.google.dev/gemma/docs/core/model_card_4", "measuredAt": "2026-04-02" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 43.4, "percentile": 1.4, "peerCount": 111, "sourceUrl": "https://ai.google.dev/gemma/docs/core/model_card_4", "measuredAt": "2026-04-02" }, { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 60, "percentile": 5.2, "peerCount": 29, "sourceUrl": "https://ai.google.dev/gemma/docs/core/model_card_4", "measuredAt": "2026-04-02" } ] }, { "modelId": "gemma-4-e4b", "name": "Gemma 4 E4B", "providerId": "google", "releaseDate": "2026-04-02", "rank": null, "score": null, "categoryScores": { "Coding": 13.6, "Reasoning": 10.8, "Agents": null }, "coverage": 3, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "livecodebench-6", "benchmark": "LiveCodeBench v6", "category": "Coding", "rawScore": 52, "percentile": 13.6, "peerCount": 11, "sourceUrl": "https://ai.google.dev/gemma/docs/core/model_card_4", "measuredAt": "2026-04-02" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 58.6, "percentile": 9.5, "peerCount": 111, "sourceUrl": "https://ai.google.dev/gemma/docs/core/model_card_4", "measuredAt": "2026-04-02" }, { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 69.4, "percentile": 12.1, "peerCount": 29, "sourceUrl": "https://ai.google.dev/gemma/docs/core/model_card_4", "measuredAt": "2026-04-02" } ] }, { "modelId": "qwen3-6-plus", "name": "Qwen3.6-Plus", "providerId": "qwen", "releaseDate": "2026-04-02", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 47.4, "Agents": null }, "coverage": 2, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 88.4, "percentile": 61.7, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 38.2, "percentile": 33.1, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "minimax-m2-7", "name": "MiniMax M2.7", "providerId": "minimax", "releaseDate": "2026-03-18", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "mistral-small-4", "name": "Mistral Small 4", "providerId": "mistral", "releaseDate": "2026-03-16", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 28.3, "Agents": null }, "coverage": 2, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 71.2, "percentile": 20.3, "peerCount": 111, "sourceUrl": "https://huggingface.co/mistralai/Mistral-Small-4-119B-2603", "measuredAt": "2026-03-16" }, { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 78, "percentile": 36.2, "peerCount": 29, "sourceUrl": "https://huggingface.co/mistralai/Mistral-Small-4-119B-2603", "measuredAt": "2026-03-16" } ] }, { "modelId": "nvidia-nemotron-3-super-120b-a12b", "name": "NVIDIA Nemotron 3 Super 120B-A12B", "providerId": "nvidia", "releaseDate": "2026-03-11", "rank": null, "score": null, "categoryScores": { "Coding": 24, "Reasoning": 41, "Agents": null }, "coverage": 5, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 60.47, "percentile": 24, "peerCount": 48, "sourceUrl": "https://build.nvidia.com/nvidia/nemotron-3-super-120b-a12b/modelcard", "measuredAt": "2026-03-11" }, { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 83.73, "percentile": 60.3, "peerCount": 29, "sourceUrl": "https://build.nvidia.com/nvidia/nemotron-3-super-120b-a12b/modelcard", "measuredAt": "2026-03-11" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 90.21, "percentile": 63.7, "peerCount": 40, "sourceUrl": "https://build.nvidia.com/nvidia/nemotron-3-super-120b-a12b/modelcard", "measuredAt": "2026-03-11" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 18.26, "percentile": 29.3, "peerCount": 29, "sourceUrl": "https://build.nvidia.com/nvidia/nemotron-3-super-120b-a12b/modelcard", "measuredAt": "2026-03-11" }, { "axisId": "hle-with-tools", "benchmark": "Humanity's Last Exam (with tools)", "category": "Reasoning", "rawScore": 22.82, "percentile": 10.7, "peerCount": 14, "sourceUrl": "https://build.nvidia.com/nvidia/nemotron-3-super-120b-a12b/modelcard", "measuredAt": "2026-03-11" } ] }, { "modelId": "grok-4-20-reasoning", "name": "Grok 4.20 Reasoning", "providerId": "xai", "releaseDate": "2026-03-09", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 51.8, "Agents": null }, "coverage": 2, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 89.3, "percentile": 66.2, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 44.9, "percentile": 37.3, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "grok-4-20-multi-agent", "name": "Grok 4.20 Multi-Agent", "providerId": "xai", "releaseDate": "2026-03-09", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "grok-4-20-non-reasoning", "name": "Grok 4.20 Non-Reasoning", "providerId": "xai", "releaseDate": "2026-03-09", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "gpt-5-4-pro", "name": "GPT-5.4 Pro", "providerId": "openai", "releaseDate": "2026-03-05", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 84.9, "Agents": 90 }, "coverage": 5, "categoryCoverage": 2, "missingCategories": [ "Coding" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 94.4, "percentile": 95.9, "peerCount": 111, "sourceUrl": "https://openai.com/index/introducing-gpt-5-4/", "measuredAt": "2026-03-05" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 42.7, "percentile": 74.1, "peerCount": 29, "sourceUrl": "https://openai.com/index/introducing-gpt-5-4/", "measuredAt": "2026-03-05" }, { "axisId": "hle-with-tools", "benchmark": "Humanity's Last Exam (with tools)", "category": "Reasoning", "rawScore": 58.7, "percentile": 82.1, "peerCount": 14, "sourceUrl": "https://openai.com/index/introducing-gpt-5-4/", "measuredAt": "2026-03-05" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 82.5, "percentile": 87.3, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" }, { "axisId": "browsecomp", "benchmark": "BrowseComp", "category": "Agents", "rawScore": 89.3, "percentile": 90, "peerCount": 15, "sourceUrl": "https://openai.com/index/introducing-gpt-5-4/", "measuredAt": "2026-03-05" } ] }, { "modelId": "gemini-3-1-flash-lite", "name": "Gemini 3.1 Flash-Lite", "providerId": "google", "releaseDate": "2026-03-03", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 33.5, "Agents": null }, "coverage": 3, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 86.9, "percentile": 53.6, "peerCount": 111, "sourceUrl": "https://deepmind.google/models/model-cards/gemini-3-1-flash-lite/", "measuredAt": "2026-03-03" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 16, "percentile": 22.4, "peerCount": 29, "sourceUrl": "https://deepmind.google/models/model-cards/gemini-3-1-flash-lite/", "measuredAt": "2026-03-03" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 27.7, "percentile": 24.6, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "qwen3-5-flash", "name": "Qwen3.5-Flash", "providerId": "qwen", "releaseDate": "2026-02-23", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 21.6, "Agents": null }, "coverage": 2, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 82.3, "percentile": 35.6, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 9.5, "percentile": 7.6, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "gemini-3-1-pro-preview", "name": "Gemini 3.1 Pro Preview", "providerId": "google", "releaseDate": "2026-02-19", "rank": null, "score": null, "categoryScores": { "Coding": 25, "Reasoning": 77.6, "Agents": null }, "coverage": 3, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-pro", "benchmark": "SWE-bench Pro", "category": "Coding", "rawScore": 54.2, "percentile": 25, "peerCount": 22, "sourceUrl": "https://openai.com/index/gpt-5-6/", "measuredAt": "2026-07-09" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 94.3, "percentile": 95, "peerCount": 111, "sourceUrl": "https://openai.com/index/gpt-5-6/", "measuredAt": "2026-07-09" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 59.6, "percentile": 60.2, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "qwen3-5-397b-a17b", "name": "Qwen3.5 397B-A17B", "providerId": "qwen", "releaseDate": "2026-02-15", "rank": null, "score": null, "categoryScores": { "Coding": 69.8, "Reasoning": 60.9, "Agents": null }, "coverage": 4, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 76.4, "percentile": 69.8, "peerCount": 48, "sourceUrl": "https://qwen.ai/blog?id=qwen3.5", "measuredAt": "2026-02-15" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 88.4, "percentile": 61.7, "peerCount": 111, "sourceUrl": "https://qwen.ai/blog?id=qwen3.6-27b", "measuredAt": "2026-04-21" }, { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 87.8, "percentile": 94.8, "peerCount": 29, "sourceUrl": "https://qwen.ai/blog?id=qwen3.5", "measuredAt": "2026-02-15" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 29.5, "percentile": 26.3, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "qwen3-5-plus", "name": "Qwen3.5-Plus", "providerId": "qwen", "releaseDate": "2026-02-15", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 46.4, "Agents": null }, "coverage": 1, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 84.8, "percentile": 46.4, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" } ] }, { "modelId": "minimax-m2-5", "name": "MiniMax M2.5", "providerId": "minimax", "releaseDate": "2026-02-12", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "ernie-5-0", "name": "ERNIE 5.0", "providerId": "baidu", "releaseDate": "2026-02-06", "rank": null, "score": null, "categoryScores": { "Coding": 40.9, "Reasoning": 54.8, "Agents": null }, "coverage": 4, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "livecodebench-6", "benchmark": "LiveCodeBench v6", "category": "Coding", "rawScore": 76.21, "percentile": 40.9, "peerCount": 11, "sourceUrl": "https://ernie.baidu.com/blog/posts/ernie5.0/", "measuredAt": "2026-02-06" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 86.36, "percentile": 51.8, "peerCount": 111, "sourceUrl": "https://ernie.baidu.com/blog/posts/ernie5.0/", "measuredAt": "2026-02-06" }, { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 83.8, "percentile": 63.8, "peerCount": 29, "sourceUrl": "https://ernie.baidu.com/blog/posts/ernie5.0/", "measuredAt": "2026-02-06" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 89.06, "percentile": 48.8, "peerCount": 40, "sourceUrl": "https://ernie.baidu.com/blog/posts/ernie5.0/", "measuredAt": "2026-02-06" } ] }, { "modelId": "qwen3-coder-next", "name": "Qwen3-Coder-Next", "providerId": "qwen", "releaseDate": "2026-02-03", "rank": null, "score": null, "categoryScores": { "Coding": 25.8, "Reasoning": null, "Agents": null }, "coverage": 2, "categoryCoverage": 1, "missingCategories": [ "Reasoning", "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 70.6, "percentile": 44.8, "peerCount": 48, "sourceUrl": "https://huggingface.co/Qwen/Qwen3-Coder-Next", "measuredAt": "2026-02-03" }, { "axisId": "swe-pro", "benchmark": "SWE-bench Pro", "category": "Coding", "rawScore": 44.3, "percentile": 6.8, "peerCount": 22, "sourceUrl": "https://huggingface.co/Qwen/Qwen3-Coder-Next", "measuredAt": "2026-02-03" } ] }, { "modelId": "glm-4-7", "name": "GLM-4.7", "providerId": "zai", "releaseDate": "2025-12-22", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 41, "Agents": null }, "coverage": 1, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 83.3, "percentile": 41, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" } ] }, { "modelId": "gemini-3-flash-preview", "name": "Gemini 3 Flash Preview", "providerId": "google", "releaseDate": "2025-12-17", "rank": null, "score": null, "categoryScores": { "Coding": 85.4, "Reasoning": 64.6, "Agents": null }, "coverage": 5, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 78, "percentile": 85.4, "peerCount": 48, "sourceUrl": "https://deepmind.google/technologies/gemini/flash/", "measuredAt": "2025-12-17" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 90.4, "percentile": 70.3, "peerCount": 111, "sourceUrl": "https://blog.google/products-and-platforms/products/gemini/gemini-3-flash/", "measuredAt": "2025-12-17" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 95.2, "percentile": 83.8, "peerCount": 40, "sourceUrl": "https://deepmind.google/models/model-cards/gemini-3-flash/", "measuredAt": "2025-12-17" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 33.7, "percentile": 60.3, "peerCount": 29, "sourceUrl": "https://blog.google/products-and-platforms/products/gemini/gemini-3-flash/", "measuredAt": "2025-12-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 51.2, "percentile": 44.1, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "nvidia-nemotron-3-nano-30b-a3b-bf16", "name": "NVIDIA Nemotron 3 Nano 30B-A3B BF16", "providerId": "nvidia", "releaseDate": "2025-12-15", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 27.5, "Agents": null }, "coverage": 4, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 78.3, "percentile": 39.7, "peerCount": 29, "sourceUrl": "https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16", "measuredAt": "2025-12-15" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 89.1, "percentile": 51.2, "peerCount": 40, "sourceUrl": "https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16", "measuredAt": "2025-12-15" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 10.6, "percentile": 15.5, "peerCount": 29, "sourceUrl": "https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16", "measuredAt": "2025-12-15" }, { "axisId": "hle-with-tools", "benchmark": "Humanity's Last Exam (with tools)", "category": "Reasoning", "rawScore": 15.5, "percentile": 3.6, "peerCount": 14, "sourceUrl": "https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16", "measuredAt": "2025-12-15" } ] }, { "modelId": "gpt-5-2-pro", "name": "GPT-5.2 Pro", "providerId": "openai", "releaseDate": "2025-12-11", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 72, "Agents": 43.3 }, "coverage": 6, "categoryCoverage": 2, "missingCategories": [ "Coding" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 93.2, "percentile": 87.8, "peerCount": 111, "sourceUrl": "https://openai.com/index/introducing-gpt-5-2/", "measuredAt": "2025-12-11" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 100, "percentile": 97.5, "peerCount": 40, "sourceUrl": "https://openai.com/index/introducing-gpt-5-2/", "measuredAt": "2025-12-11" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 36.6, "percentile": 63.8, "peerCount": 29, "sourceUrl": "https://openai.com/index/introducing-gpt-5-2/", "measuredAt": "2025-12-11" }, { "axisId": "hle-with-tools", "benchmark": "Humanity's Last Exam (with tools)", "category": "Reasoning", "rawScore": 50, "percentile": 32.1, "peerCount": 14, "sourceUrl": "https://openai.com/index/introducing-gpt-5-2/", "measuredAt": "2025-12-11" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 74, "percentile": 78.8, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" }, { "axisId": "browsecomp", "benchmark": "BrowseComp", "category": "Agents", "rawScore": 77.9, "percentile": 43.3, "peerCount": 15, "sourceUrl": "https://openai.com/index/introducing-gpt-5-2/", "measuredAt": "2025-12-11" } ] }, { "modelId": "gpt-5-2", "name": "GPT-5.2", "providerId": "openai", "releaseDate": "2025-12-11", "rank": null, "score": null, "categoryScores": { "Coding": 62.4, "Reasoning": 82.7, "Agents": null }, "coverage": 5, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 80, "percentile": 90.6, "peerCount": 48, "sourceUrl": "https://openai.com/index/introducing-gpt-5-2/", "measuredAt": "2025-12-11" }, { "axisId": "swe-pro", "benchmark": "SWE-bench Pro", "category": "Coding", "rawScore": 55.6, "percentile": 34.1, "peerCount": 22, "sourceUrl": "https://openai.com/index/introducing-gpt-5-2/", "measuredAt": "2025-12-11" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 92.4, "percentile": 82, "peerCount": 111, "sourceUrl": "https://openai.com/index/introducing-gpt-5-2/", "measuredAt": "2025-12-11" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 100, "percentile": 97.5, "peerCount": 40, "sourceUrl": "https://openai.com/index/introducing-gpt-5-2/", "measuredAt": "2025-12-11" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 67.4, "percentile": 68.6, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "amazon-nova-2-lite", "name": "Amazon Nova 2 Lite", "providerId": "amazon", "releaseDate": "2025-12-02", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 47.7, "Agents": 12.2 }, "coverage": 5, "categoryCoverage": 2, "missingCategories": [ "Coding" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 79.6, "percentile": 30.2, "peerCount": 111, "sourceUrl": "https://cdn.amazon.science/c5/3d/84514a224666b5be6de4b43ef4aa/nova-2-0-technical-report2.pdf", "measuredAt": "2025-12-02" }, { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 80.9, "percentile": 46.6, "peerCount": 29, "sourceUrl": "https://cdn.amazon.science/c5/3d/84514a224666b5be6de4b43ef4aa/nova-2-0-technical-report2.pdf", "measuredAt": "2025-12-02" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 91, "percentile": 66.3, "peerCount": 40, "sourceUrl": "https://cdn.amazon.science/c5/3d/84514a224666b5be6de4b43ef4aa/nova-2-0-technical-report2.pdf", "measuredAt": "2025-12-02" }, { "axisId": "mcp-atlas", "benchmark": "MCP Atlas", "category": "Agents", "rawScore": 24.6, "percentile": 5.6, "peerCount": 9, "sourceUrl": "https://cdn.amazon.science/c5/3d/84514a224666b5be6de4b43ef4aa/nova-2-0-technical-report2.pdf", "measuredAt": "2025-12-02" }, { "axisId": "tau2-telecom", "benchmark": "Tau2-bench Telecom", "category": "Agents", "rawScore": 76, "percentile": 18.8, "peerCount": 8, "sourceUrl": "https://cdn.amazon.science/c5/3d/84514a224666b5be6de4b43ef4aa/nova-2-0-technical-report2.pdf", "measuredAt": "2025-12-02" } ] }, { "modelId": "ministral-3-14b", "name": "Ministral 3 14B", "providerId": "mistral", "releaseDate": "2025-12-02", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 30.2, "Agents": null }, "coverage": 2, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 71.2, "percentile": 20.3, "peerCount": 111, "sourceUrl": "https://huggingface.co/mistralai/Ministral-3-14B-Instruct-2512-BF16", "measuredAt": "2025-12-02" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 85, "percentile": 40, "peerCount": 40, "sourceUrl": "https://huggingface.co/mistralai/Ministral-3-14B-Instruct-2512-BF16", "measuredAt": "2025-12-02" } ] }, { "modelId": "ministral-3-3b", "name": "Ministral 3 3B", "providerId": "mistral", "releaseDate": "2025-12-02", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 16.1, "Agents": null }, "coverage": 2, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 53.4, "percentile": 5.9, "peerCount": 111, "sourceUrl": "https://huggingface.co/mistralai/Ministral-3-3B-Instruct-2512", "measuredAt": "2025-12-02" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 72.1, "percentile": 26.3, "peerCount": 40, "sourceUrl": "https://huggingface.co/mistralai/Ministral-3-3B-Instruct-2512", "measuredAt": "2025-12-02" } ] }, { "modelId": "ministral-3-8b", "name": "Ministral 3 8B", "providerId": "mistral", "releaseDate": "2025-12-02", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 22.3, "Agents": null }, "coverage": 2, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 66.8, "percentile": 15.8, "peerCount": 111, "sourceUrl": "https://huggingface.co/mistralai/Ministral-3-8B-Instruct-2512-BF16", "measuredAt": "2025-12-02" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 78.7, "percentile": 28.7, "peerCount": 40, "sourceUrl": "https://huggingface.co/mistralai/Ministral-3-8B-Instruct-2512-BF16", "measuredAt": "2025-12-02" } ] }, { "modelId": "mistral-large-3", "name": "Mistral Large 3", "providerId": "mistral", "releaseDate": "2025-12-02", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "deepseek-v3-2", "name": "DeepSeek V3.2", "providerId": "deepseek", "releaseDate": "2025-12-01", "rank": null, "score": null, "categoryScores": { "Coding": 53.1, "Reasoning": 56.7, "Agents": null }, "coverage": 5, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 73.1, "percentile": 53.1, "peerCount": 48, "sourceUrl": "https://modelscope.cn/models/deepseek-ai/DeepSeek-V3.2/resolve/master/assets/paper.pdf", "measuredAt": "2025-12-01" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 82.4, "percentile": 37.4, "peerCount": 111, "sourceUrl": "https://modelscope.cn/models/deepseek-ai/DeepSeek-V3.2/resolve/master/assets/paper.pdf", "measuredAt": "2025-12-01" }, { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 85, "percentile": 72.4, "peerCount": 29, "sourceUrl": "https://modelscope.cn/models/deepseek-ai/DeepSeek-V3.2/resolve/master/assets/paper.pdf", "measuredAt": "2025-12-01" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 93.1, "percentile": 73.8, "peerCount": 40, "sourceUrl": "https://modelscope.cn/models/deepseek-ai/DeepSeek-V3.2/resolve/master/assets/paper.pdf", "measuredAt": "2025-12-01" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 25.1, "percentile": 43.1, "peerCount": 29, "sourceUrl": "https://modelscope.cn/models/deepseek-ai/DeepSeek-V3.2/resolve/master/assets/paper.pdf", "measuredAt": "2025-12-01" } ] }, { "modelId": "gpt-5-1", "name": "GPT-5.1", "providerId": "openai", "releaseDate": "2025-11-13", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 66.7, "Agents": null }, "coverage": 2, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 87.6, "percentile": 54.5, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 94.2, "percentile": 78.8, "peerCount": 40, "sourceUrl": "https://matharena.ai/?comp=aime--aime_2025", "measuredAt": "2026-09-17" } ] }, { "modelId": "claude-haiku-4-5", "name": "Claude Haiku 4.5", "providerId": "anthropic", "releaseDate": "2025-10-15", "rank": null, "score": null, "categoryScores": { "Coding": 55.2, "Reasoning": 20.3, "Agents": null }, "coverage": 2, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 73.3, "percentile": 55.2, "peerCount": 48, "sourceUrl": "https://www.anthropic.com/news/claude-haiku-4-5", "measuredAt": "2025-10-15" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 71.2, "percentile": 20.3, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" } ] }, { "modelId": "gpt-5-pro", "name": "GPT-5 Pro", "providerId": "openai", "releaseDate": "2025-10-06", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 56.3, "Agents": null }, "coverage": 2, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 88.4, "percentile": 61.7, "peerCount": 111, "sourceUrl": "https://openai.com/index/introducing-gpt-5/", "measuredAt": "2025-08-07" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 55.8, "percentile": 50.8, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "qwen3-max", "name": "Qwen3-Max", "providerId": "qwen", "releaseDate": "2025-09-23", "rank": null, "score": null, "categoryScores": { "Coding": 35.4, "Reasoning": 19.2, "Agents": null }, "coverage": 3, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 69.6, "percentile": 35.4, "peerCount": 48, "sourceUrl": "https://qwen.ai/blog?from=research.latest-advancements-list&id=241398b9cd6353de490b0f82806c7848c5d2777d", "measuredAt": "2025-09-24" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 72.6, "percentile": 23.9, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 18.9, "percentile": 14.4, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "grok-4-fast-reasoning", "name": "Grok 4 Fast Reasoning", "providerId": "xai", "releaseDate": "2025-09-19", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 51.9, "Agents": 16.7 }, "coverage": 4, "categoryCoverage": 2, "missingCategories": [ "Coding" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 85.7, "percentile": 48.2, "peerCount": 111, "sourceUrl": "https://x.ai/news/grok-4-fast", "measuredAt": "2025-09-19" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 92, "percentile": 71.3, "peerCount": 40, "sourceUrl": "https://x.ai/news/grok-4-fast", "measuredAt": "2025-09-19" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 20, "percentile": 36.2, "peerCount": 29, "sourceUrl": "https://x.ai/news/grok-4-fast", "measuredAt": "2025-09-19" }, { "axisId": "browsecomp", "benchmark": "BrowseComp", "category": "Agents", "rawScore": 44.9, "percentile": 16.7, "peerCount": 15, "sourceUrl": "https://x.ai/news/grok-4-fast", "measuredAt": "2025-09-19" } ] }, { "modelId": "longcat-flash-chat", "name": "LongCat Flash Chat", "providerId": "meituan", "releaseDate": "2025-09-01", "rank": null, "score": null, "categoryScores": { "Coding": 21.9, "Reasoning": 31.8, "Agents": null }, "coverage": 4, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 60.4, "percentile": 21.9, "peerCount": 48, "sourceUrl": "https://github.com/meituan-longcat/LongCat-Flash-Chat", "measuredAt": "2025-09-01" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 73.23, "percentile": 24.8, "peerCount": 111, "sourceUrl": "https://github.com/meituan-longcat/LongCat-Flash-Chat", "measuredAt": "2025-09-01" }, { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 82.68, "percentile": 56.9, "peerCount": 29, "sourceUrl": "https://github.com/meituan-longcat/LongCat-Flash-Chat", "measuredAt": "2025-09-01" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 61.25, "percentile": 13.8, "peerCount": 40, "sourceUrl": "https://github.com/meituan-longcat/LongCat-Flash-Chat", "measuredAt": "2025-09-01" } ] }, { "modelId": "command-a-translate", "name": "Command A Translate", "providerId": "cohere", "releaseDate": "2025-08-28", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "command-a-reasoning", "name": "Command A Reasoning", "providerId": "cohere", "releaseDate": "2025-08-21", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 11.3, "Agents": 6.3 }, "coverage": 2, "categoryCoverage": 2, "missingCategories": [ "Coding" ], "components": [ { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 57, "percentile": 11.3, "peerCount": 40, "sourceUrl": "https://cohere.com/blog/command-a-plus", "measuredAt": "2026-05-20" }, { "axisId": "tau2-telecom", "benchmark": "Tau2-bench Telecom", "category": "Agents", "rawScore": 37, "percentile": 6.3, "peerCount": 8, "sourceUrl": "https://cohere.com/blog/command-a-plus", "measuredAt": "2026-05-20" } ] }, { "modelId": "gpt-5", "name": "GPT-5", "providerId": "openai", "releaseDate": "2025-08-07", "rank": null, "score": null, "categoryScores": { "Coding": 65.6, "Reasoning": 59.9, "Agents": null }, "coverage": 4, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 74.9, "percentile": 65.6, "peerCount": 48, "sourceUrl": "https://openai.com/index/introducing-gpt-5/", "measuredAt": "2025-08-07" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 86.2, "percentile": 50, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 94.6, "percentile": 81.3, "peerCount": 40, "sourceUrl": "https://openai.com/index/introducing-gpt-5/", "measuredAt": "2025-08-07" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 55.4, "percentile": 48.3, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "gpt-5-mini", "name": "GPT-5 mini", "providerId": "openai", "releaseDate": "2025-08-07", "rank": null, "score": null, "categoryScores": { "Coding": 46.9, "Reasoning": 39.4, "Agents": null }, "coverage": 4, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 71, "percentile": 46.9, "peerCount": 48, "sourceUrl": "https://openai.com/index/introducing-gpt-5-for-developers/", "measuredAt": "2025-08-07" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 81.6, "percentile": 32.9, "peerCount": 111, "sourceUrl": "https://openai.com/index/introducing-gpt-5-4-mini-and-nano/", "measuredAt": "2026-03-17" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 87.5, "percentile": 43.8, "peerCount": 40, "sourceUrl": "https://matharena.ai/?comp=aime--aime_2025", "measuredAt": "2026-09-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 46.7, "percentile": 41.5, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "gpt-5-nano", "name": "GPT-5 nano", "providerId": "openai", "releaseDate": "2025-08-07", "rank": null, "score": null, "categoryScores": { "Coding": 17.7, "Reasoning": 24.8, "Agents": null }, "coverage": 4, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 54.7, "percentile": 17.7, "peerCount": 48, "sourceUrl": "https://openai.com/index/introducing-gpt-5-for-developers/", "measuredAt": "2025-08-07" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 69.4, "percentile": 16.7, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 85, "percentile": 40, "peerCount": 40, "sourceUrl": "https://matharena.ai/?comp=aime--aime_2025", "measuredAt": "2026-09-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 20, "percentile": 17.8, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "claude-opus-4-1", "name": "Claude Opus 4.1", "providerId": "anthropic", "releaseDate": "2025-08-05", "rank": null, "score": null, "categoryScores": { "Coding": 63.5, "Reasoning": 18.4, "Agents": null }, "coverage": 3, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 74.5, "percentile": 63.5, "peerCount": 48, "sourceUrl": "https://www.anthropic.com/news/claude-opus-4-1", "measuredAt": "2025-08-05" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 77.3, "percentile": 27.5, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 12.6, "percentile": 9.3, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "gpt-oss-120b", "name": "gpt-oss-120b", "providerId": "openai", "releaseDate": "2025-08-05", "rank": null, "score": null, "categoryScores": { "Coding": 28.1, "Reasoning": 46, "Agents": null }, "coverage": 3, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 62.4, "percentile": 28.1, "peerCount": 48, "sourceUrl": "https://cdn.openai.com/pdf/419b6906-9da6-406c-a19d-1bb078ac7637/oai_gpt-oss_model_card.pdf", "measuredAt": "2025-08-05" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 80.1, "percentile": 32, "peerCount": 111, "sourceUrl": "https://cdn.openai.com/pdf/419b6906-9da6-406c-a19d-1bb078ac7637/oai_gpt-oss_model_card.pdf", "measuredAt": "2025-08-05" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 90, "percentile": 60, "peerCount": 40, "sourceUrl": "https://matharena.ai/?comp=aime--aime_2025", "measuredAt": "2026-09-17" } ] }, { "modelId": "gpt-oss-20b", "name": "gpt-oss-20b", "providerId": "openai", "releaseDate": "2025-08-05", "rank": null, "score": null, "categoryScores": { "Coding": 26, "Reasoning": 38.8, "Agents": null }, "coverage": 3, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 60.7, "percentile": 26, "peerCount": 48, "sourceUrl": "https://cdn.openai.com/pdf/419b6906-9da6-406c-a19d-1bb078ac7637/oai_gpt-oss_model_card.pdf", "measuredAt": "2025-08-05" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 71.5, "percentile": 22.5, "peerCount": 111, "sourceUrl": "https://cdn.openai.com/pdf/419b6906-9da6-406c-a19d-1bb078ac7637/oai_gpt-oss_model_card.pdf", "measuredAt": "2025-08-05" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 89.2, "percentile": 55, "peerCount": 40, "sourceUrl": "https://matharena.ai/?comp=aime--aime_2025", "measuredAt": "2026-09-17" } ] }, { "modelId": "command-a-vision", "name": "Command A Vision", "providerId": "cohere", "releaseDate": "2025-07-31", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "gemini-2-5-flash-lite", "name": "Gemini 2.5 Flash-Lite", "providerId": "google", "releaseDate": "2025-07-22", "rank": null, "score": null, "categoryScores": { "Coding": 1, "Reasoning": 13, "Agents": null }, "coverage": 4, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 27.6, "percentile": 1, "peerCount": 48, "sourceUrl": "https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-5-Flash-Lite-Model-Card.pdf", "measuredAt": "2025-06-17" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 66.7, "percentile": 14.9, "peerCount": 111, "sourceUrl": "https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-5-Flash-Lite-Model-Card.pdf", "measuredAt": "2025-06-17" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 63.1, "percentile": 18.8, "peerCount": 40, "sourceUrl": "https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-5-Flash-Lite-Model-Card.pdf", "measuredAt": "2025-06-17" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 6.9, "percentile": 5.2, "peerCount": 29, "sourceUrl": "https://storage.googleapis.com/deepmind-media/Model-Cards/Gemini-2-5-Flash-Lite-Model-Card.pdf", "measuredAt": "2025-06-17" } ] }, { "modelId": "qwen3-coder-plus", "name": "Qwen3-Coder-Plus", "providerId": "qwen", "releaseDate": "2025-07-22", "rank": null, "score": null, "categoryScores": { "Coding": 35.4, "Reasoning": null, "Agents": null }, "coverage": 1, "categoryCoverage": 1, "missingCategories": [ "Reasoning", "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 69.6, "percentile": 35.4, "peerCount": 48, "sourceUrl": "https://www.linkedin.com/posts/qwen_were-excited-to-announce-the-upgrade-of-activity-7376348152515260416-b4oF", "measuredAt": "2025-09-23" } ] }, { "modelId": "gemini-2-5-pro", "name": "Gemini 2.5 Pro", "providerId": "google", "releaseDate": "2025-06-17", "rank": null, "score": null, "categoryScores": { "Coding": 19.8, "Reasoning": 40, "Agents": null }, "coverage": 5, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 59.6, "percentile": 19.8, "peerCount": 48, "sourceUrl": "https://storage.googleapis.com/deepmind-media/gemini/gemini_v2_5_report.pdf", "measuredAt": "2025-06-17" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 86.4, "percentile": 52.7, "peerCount": 111, "sourceUrl": "https://storage.googleapis.com/deepmind-media/gemini/gemini_v2_5_report.pdf", "measuredAt": "2025-06-17" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 88, "percentile": 46.3, "peerCount": 40, "sourceUrl": "https://storage.googleapis.com/deepmind-media/gemini/gemini_v2_5_report.pdf", "measuredAt": "2025-06-17" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 21.6, "percentile": 39.7, "peerCount": 29, "sourceUrl": "https://storage.googleapis.com/deepmind-media/gemini/gemini_v2_5_report.pdf", "measuredAt": "2025-06-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 24.6, "percentile": 21.2, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "gemini-2-5-flash", "name": "Gemini 2.5 Flash", "providerId": "google", "releaseDate": "2025-06-17", "rank": null, "score": null, "categoryScores": { "Coding": 7.3, "Reasoning": 27.2, "Agents": null }, "coverage": 4, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 48.9, "percentile": 7.3, "peerCount": 48, "sourceUrl": "https://storage.googleapis.com/deepmind-media/gemini/gemini_v2_5_report.pdf", "measuredAt": "2025-06-17" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 82.8, "percentile": 38.7, "peerCount": 111, "sourceUrl": "https://storage.googleapis.com/deepmind-media/gemini/gemini_v2_5_report.pdf", "measuredAt": "2025-06-17" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 72, "percentile": 23.8, "peerCount": 40, "sourceUrl": "https://storage.googleapis.com/deepmind-media/gemini/gemini_v2_5_report.pdf", "measuredAt": "2025-06-17" }, { "axisId": "hle-no-tools", "benchmark": "Humanity's Last Exam (no tools)", "category": "Reasoning", "rawScore": 11, "percentile": 19, "peerCount": 29, "sourceUrl": "https://storage.googleapis.com/deepmind-media/gemini/gemini_v2_5_report.pdf", "measuredAt": "2025-06-17" } ] }, { "modelId": "o3-pro", "name": "o3-pro", "providerId": "openai", "releaseDate": "2025-06-10", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "claude-opus-4", "name": "Claude Opus 4", "providerId": "anthropic", "releaseDate": "2025-05-22", "rank": null, "score": null, "categoryScores": { "Coding": 49, "Reasoning": 26.6, "Agents": null }, "coverage": 2, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 72.5, "percentile": 49, "peerCount": 48, "sourceUrl": "https://www.anthropic.com/news/claude-4", "measuredAt": "2025-05-22" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 76.3, "percentile": 26.6, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" } ] }, { "modelId": "claude-sonnet-4", "name": "Claude Sonnet 4", "providerId": "anthropic", "releaseDate": "2025-05-22", "rank": null, "score": null, "categoryScores": { "Coding": 51, "Reasoning": 29.3, "Agents": null }, "coverage": 2, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 72.7, "percentile": 51, "peerCount": 48, "sourceUrl": "https://www.anthropic.com/news/claude-4", "measuredAt": "2025-05-22" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 79.2, "percentile": 29.3, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" } ] }, { "modelId": "phi-4-reasoning", "name": "Phi-4 Reasoning", "providerId": "microsoft", "releaseDate": "2025-04-30", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 17.8, "Agents": null }, "coverage": 3, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 65.8, "percentile": 13.1, "peerCount": 111, "sourceUrl": "https://huggingface.co/microsoft/Phi-4-reasoning", "measuredAt": "2025-04-30" }, { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 74.3, "percentile": 24.1, "peerCount": 29, "sourceUrl": "https://huggingface.co/microsoft/Phi-4-reasoning", "measuredAt": "2025-04-30" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 62.9, "percentile": 16.3, "peerCount": 40, "sourceUrl": "https://huggingface.co/microsoft/Phi-4-reasoning", "measuredAt": "2025-04-30" } ] }, { "modelId": "amazon-nova-premier", "name": "Amazon Nova Premier", "providerId": "amazon", "releaseDate": "2025-04-30", "rank": null, "score": null, "categoryScores": { "Coding": 5.2, "Reasoning": null, "Agents": null }, "coverage": 1, "categoryCoverage": 1, "missingCategories": [ "Reasoning", "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 42.4, "percentile": 5.2, "peerCount": 48, "sourceUrl": "https://aws.amazon.com/about-aws/whats-new/2025/04/amazon-nova-premier-complex-tasks-model-distillation/", "measuredAt": "2025-04-30" } ] }, { "modelId": "phi-4-mini-reasoning", "name": "Phi-4 Mini Reasoning", "providerId": "microsoft", "releaseDate": "2025-04-30", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 5, "Agents": null }, "coverage": 1, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 52, "percentile": 5, "peerCount": 111, "sourceUrl": "https://huggingface.co/microsoft/Phi-4-mini-reasoning", "measuredAt": "2025-04-30" } ] }, { "modelId": "qwen3-235b-a22b", "name": "Qwen3 235B-A22B", "providerId": "qwen", "releaseDate": "2025-04-29", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 24.9, "Agents": null }, "coverage": 2, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 70.7, "percentile": 18.5, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 80.8, "percentile": 31.3, "peerCount": 40, "sourceUrl": "https://matharena.ai/?comp=aime--aime_2025", "measuredAt": "2026-09-17" } ] }, { "modelId": "o3", "name": "o3", "providerId": "openai", "releaseDate": "2025-04-16", "rank": null, "score": null, "categoryScores": { "Coding": 32.3, "Reasoning": 38.9, "Agents": null }, "coverage": 4, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 69.1, "percentile": 32.3, "peerCount": 48, "sourceUrl": "https://openai.com/index/introducing-gpt-5-for-developers/", "measuredAt": "2025-08-07" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 81.8, "percentile": 33.8, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 89.2, "percentile": 55, "peerCount": 40, "sourceUrl": "https://matharena.ai/?comp=aime--aime_2025", "measuredAt": "2026-09-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 33.3, "percentile": 28, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "gpt-4-1", "name": "GPT-4.1", "providerId": "openai", "releaseDate": "2025-04-14", "rank": null, "score": null, "categoryScores": { "Coding": 15.6, "Reasoning": 9.1, "Agents": null }, "coverage": 3, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 54.6, "percentile": 15.6, "peerCount": 48, "sourceUrl": "https://openai.com/index/gpt-4-1/", "measuredAt": "2025-04-14" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 66.3, "percentile": 14, "peerCount": 111, "sourceUrl": "https://openai.com/index/gpt-4-1/", "measuredAt": "2025-04-14" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 6, "percentile": 4.2, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "gpt-4-1-mini", "name": "GPT-4.1 mini", "providerId": "openai", "releaseDate": "2025-04-14", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 9.1, "Agents": null }, "coverage": 2, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 65, "percentile": 12.2, "peerCount": 111, "sourceUrl": "https://openai.com/index/gpt-4-1/", "measuredAt": "2025-04-14" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 6.7, "percentile": 5.9, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "gpt-4-1-nano", "name": "GPT-4.1 nano", "providerId": "openai", "releaseDate": "2025-04-14", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 3.2, "Agents": null }, "coverage": 1, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 50.3, "percentile": 3.2, "peerCount": 111, "sourceUrl": "https://openai.com/index/gpt-4-1/", "measuredAt": "2025-04-14" } ] }, { "modelId": "glm-4-32b-0414-128k", "name": "GLM-4 32B 0414 128K", "providerId": "zai", "releaseDate": "2025-04-14", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "llama-4-maverick", "name": "Llama 4 Maverick", "providerId": "meta", "releaseDate": "2025-04-05", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 30.4, "Agents": null }, "coverage": 2, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 69.8, "percentile": 17.6, "peerCount": 111, "sourceUrl": "https://huggingface.co/meta-llama/Llama-4-Maverick-17B-128E-Instruct", "measuredAt": "2025-04-05" }, { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 80.5, "percentile": 43.1, "peerCount": 29, "sourceUrl": "https://huggingface.co/meta-llama/Llama-4-Maverick-17B-128E-Instruct", "measuredAt": "2025-04-05" } ] }, { "modelId": "llama-4-scout", "name": "Llama 4 Scout", "providerId": "meta", "releaseDate": "2025-04-05", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 16.4, "Agents": null }, "coverage": 2, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 57.2, "percentile": 8.6, "peerCount": 111, "sourceUrl": "https://huggingface.co/meta-llama/Llama-4-Maverick-17B-128E-Instruct", "measuredAt": "2025-04-05" }, { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 74.3, "percentile": 24.1, "peerCount": 29, "sourceUrl": "https://huggingface.co/meta-llama/Llama-4-Maverick-17B-128E-Instruct", "measuredAt": "2025-04-05" } ] }, { "modelId": "command-a", "name": "Command A", "providerId": "cohere", "releaseDate": "2025-03-13", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 15.5, "Agents": null }, "coverage": 1, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 69.6, "percentile": 15.5, "peerCount": 29, "sourceUrl": "https://cohere.com/research/papers/command-a-technical-report.pdf", "measuredAt": "2025-03-13" } ] }, { "modelId": "phi-4-multimodal-instruct", "name": "Phi-4 Multimodal Instruct", "providerId": "microsoft", "releaseDate": "2025-02-26", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "claude-sonnet-3-7", "name": "Claude 3.7 Sonnet", "providerId": "anthropic", "releaseDate": "2025-02-24", "rank": null, "score": null, "categoryScores": { "Coding": 40.6, "Reasoning": 20, "Agents": null }, "coverage": 3, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 70.3, "percentile": 40.6, "peerCount": 48, "sourceUrl": "https://www.anthropic.com/news/claude-3-7-sonnet", "measuredAt": "2025-02-24" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 79.7, "percentile": 31.1, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 49.2, "percentile": 8.8, "peerCount": 40, "sourceUrl": "https://matharena.ai/?comp=aime--aime_2025", "measuredAt": "2026-09-17" } ] }, { "modelId": "sonar-deep-research", "name": "Sonar Deep Research", "providerId": "perplexity", "releaseDate": "2025-02-14", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "sonar", "name": "Sonar", "providerId": "perplexity", "releaseDate": "2025-01-21", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "sonar-pro", "name": "Sonar Pro", "providerId": "perplexity", "releaseDate": "2025-01-21", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "deepseek-r1", "name": "DeepSeek R1", "providerId": "deepseek", "releaseDate": "2025-01-20", "rank": null, "score": null, "categoryScores": { "Coding": 11.5, "Reasoning": 37, "Agents": null }, "coverage": 4, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 49.2, "percentile": 11.5, "peerCount": 48, "sourceUrl": "https://github.com/deepseek-ai/DeepSeek-R1/blob/main/README.md", "measuredAt": "2025-01-20" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 71.5, "percentile": 22.5, "peerCount": 111, "sourceUrl": "https://github.com/deepseek-ai/DeepSeek-R1/blob/main/README.md", "measuredAt": "2025-01-20" }, { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 84, "percentile": 67.2, "peerCount": 29, "sourceUrl": "https://github.com/deepseek-ai/DeepSeek-R1/blob/main/README.md", "measuredAt": "2025-01-20" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 70, "percentile": 21.3, "peerCount": 40, "sourceUrl": "https://matharena.ai/?comp=aime--aime_2025", "measuredAt": "2026-09-17" } ] }, { "modelId": "codestral", "name": "Codestral", "providerId": "mistral", "releaseDate": "2025-01-13", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "deepseek-v3", "name": "DeepSeek V3", "providerId": "deepseek", "releaseDate": "2024-12-26", "rank": null, "score": null, "categoryScores": { "Coding": 3.1, "Reasoning": 15.3, "Agents": null }, "coverage": 4, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 42, "percentile": 3.1, "peerCount": 48, "sourceUrl": "https://github.com/deepseek-ai/DeepSeek-V3", "measuredAt": "2024-12-26" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 59.1, "percentile": 10.4, "peerCount": 111, "sourceUrl": "https://github.com/deepseek-ai/DeepSeek-V3", "measuredAt": "2024-12-26" }, { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 75.9, "percentile": 29.3, "peerCount": 29, "sourceUrl": "https://github.com/deepseek-ai/DeepSeek-V3", "measuredAt": "2024-12-26" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 25, "percentile": 6.3, "peerCount": 40, "sourceUrl": "https://matharena.ai/?comp=aime--aime_2025", "measuredAt": "2026-09-17" } ] }, { "modelId": "command-r7b", "name": "Command R7B", "providerId": "cohere", "releaseDate": "2024-12-16", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 1.7, "Agents": null }, "coverage": 1, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 42.4, "percentile": 1.7, "peerCount": 29, "sourceUrl": "https://cohere.com/research/papers/command-a-technical-report.pdf", "measuredAt": "2025-03-13" } ] }, { "modelId": "phi-4", "name": "Phi-4", "providerId": "microsoft", "releaseDate": "2024-12-12", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 13.4, "Agents": null }, "coverage": 2, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 56.1, "percentile": 7.7, "peerCount": 111, "sourceUrl": "https://www.microsoft.com/en-us/research/wp-content/uploads/2024/12/P4TechReport.pdf", "measuredAt": "2024-12-12" }, { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 70.4, "percentile": 19, "peerCount": 29, "sourceUrl": "https://www.microsoft.com/en-us/research/wp-content/uploads/2024/12/P4TechReport.pdf", "measuredAt": "2024-12-12" } ] }, { "modelId": "llama-3-3-70b-instruct", "name": "Llama 3.3 70B Instruct", "providerId": "meta", "releaseDate": "2024-12-06", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 6.4, "Agents": null }, "coverage": 2, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 50.5, "percentile": 4.1, "peerCount": 111, "sourceUrl": "https://huggingface.co/meta-llama/Llama-4-Maverick-17B-128E-Instruct", "measuredAt": "2025-04-05" }, { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 68.9, "percentile": 8.6, "peerCount": 29, "sourceUrl": "https://huggingface.co/meta-llama/Llama-4-Maverick-17B-128E-Instruct", "measuredAt": "2025-04-05" } ] }, { "modelId": "amazon-nova-lite", "name": "Amazon Nova Lite", "providerId": "amazon", "releaseDate": "2024-12-02", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "amazon-nova-pro", "name": "Amazon Nova Pro", "providerId": "amazon", "releaseDate": "2024-12-02", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "claude-sonnet-3-5", "name": "Claude 3.5 Sonnet", "providerId": "anthropic", "releaseDate": "2024-10-22", "rank": null, "score": null, "categoryScores": { "Coding": 9.4, "Reasoning": 4.1, "Agents": null }, "coverage": 3, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 49, "percentile": 9.4, "peerCount": 48, "sourceUrl": "https://www.anthropic.com/news/claude-3-5-sonnet", "measuredAt": "2024-10-22" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 55.3, "percentile": 6.8, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 3.3, "percentile": 1.3, "peerCount": 40, "sourceUrl": "https://matharena.ai/?comp=aime--aime_2025", "measuredAt": "2026-09-17" } ] }, { "modelId": "command-r", "name": "Command R", "providerId": "cohere", "releaseDate": "2024-08-30", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "command-r-plus", "name": "Command R+", "providerId": "cohere", "releaseDate": "2024-08-30", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "gpt-4o-mini", "name": "GPT-4o mini", "providerId": "openai", "releaseDate": "2024-07-18", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 1.5, "Agents": null }, "coverage": 2, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 40.2, "percentile": 0.5, "peerCount": 111, "sourceUrl": "https://openai.com/index/gpt-4-1/", "measuredAt": "2025-04-14" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 0.7, "percentile": 2.5, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "gpt-4o", "name": "GPT-4o", "providerId": "openai", "releaseDate": "2024-05-13", "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 2.3, "Agents": null }, "coverage": 3, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 46, "percentile": 2.3, "peerCount": 111, "sourceUrl": "https://openai.com/index/gpt-4-1/", "measuredAt": "2025-04-14" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 11.7, "percentile": 3.8, "peerCount": 40, "sourceUrl": "https://matharena.ai/?comp=aime--aime_2025", "measuredAt": "2026-09-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 0.4, "percentile": 0.8, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "qwen3-6-35b-a3b", "name": "Qwen3.6 35B-A3B", "providerId": "qwen", "releaseDate": null, "rank": null, "score": null, "categoryScores": { "Coding": 57.3, "Reasoning": 47, "Agents": null }, "coverage": 4, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 73.4, "percentile": 57.3, "peerCount": 48, "sourceUrl": "https://qwen.ai/blog?id=qwen3.6-27b", "measuredAt": "2026-04-21" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 86, "percentile": 49.1, "peerCount": 111, "sourceUrl": "https://qwen.ai/blog?id=qwen3.6-27b", "measuredAt": "2026-04-21" }, { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 85.2, "percentile": 79.3, "peerCount": 29, "sourceUrl": "https://qwen.ai/blog?id=qwen3.6-27b", "measuredAt": "2026-04-21" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 17.5, "percentile": 12.7, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "glm-5", "name": "GLM-5", "providerId": "zai", "releaseDate": null, "rank": null, "score": null, "categoryScores": { "Coding": 82.3, "Reasoning": 71.1, "Agents": null }, "coverage": 3, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 77.8, "percentile": 82.3, "peerCount": 48, "sourceUrl": "https://docs.z.ai/guides/llm/glm-5", "measuredAt": "2026-02-11" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 87.8, "percentile": 55.9, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 96.7, "percentile": 86.3, "peerCount": 40, "sourceUrl": "https://matharena.ai/?comp=aime--aime_2025", "measuredAt": "2026-09-17" } ] }, { "modelId": "qwen3-5-27b", "name": "Qwen3.5 27B", "providerId": "qwen", "releaseDate": null, "rank": null, "score": null, "categoryScores": { "Coding": 67.7, "Reasoning": 65.9, "Agents": null }, "coverage": 3, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-verified", "benchmark": "SWE-bench Verified", "category": "Coding", "rawScore": 75, "percentile": 67.7, "peerCount": 48, "sourceUrl": "https://qwen.ai/blog?id=qwen3.6-27b", "measuredAt": "2026-04-21" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 85.5, "percentile": 47.3, "peerCount": 111, "sourceUrl": "https://qwen.ai/blog?id=qwen3.6-27b", "measuredAt": "2026-04-21" }, { "axisId": "mmlu-pro", "benchmark": "MMLU-Pro", "category": "Reasoning", "rawScore": 86.1, "percentile": 84.5, "peerCount": 29, "sourceUrl": "https://qwen.ai/blog?id=qwen3.6-27b", "measuredAt": "2026-04-21" } ] }, { "modelId": "qwen3-8-flash-next", "name": "Qwen3.8-Flash-Next", "providerId": "qwen", "releaseDate": null, "rank": null, "score": null, "categoryScores": { "Coding": 78.5, "Reasoning": 78.8, "Agents": null }, "coverage": 3, "categoryCoverage": 2, "missingCategories": [ "Agents" ], "components": [ { "axisId": "swe-pro", "benchmark": "SWE-bench Pro", "category": "Coding", "rawScore": 62.5, "percentile": 70.5, "peerCount": 22, "sourceUrl": "https://huggingface.co/Qwen/Qwen3.8-Flash-Next", "measuredAt": "2026-09-16" }, { "axisId": "livecodebench-6", "benchmark": "LiveCodeBench v6", "category": "Coding", "rawScore": 91.9, "percentile": 86.4, "peerCount": 11, "sourceUrl": "https://huggingface.co/Qwen/Qwen3.8-Flash-Next", "measuredAt": "2026-09-16" }, { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 91.7, "percentile": 78.8, "peerCount": 111, "sourceUrl": "https://huggingface.co/Qwen/Qwen3.8-Flash-Next", "measuredAt": "2026-09-16" } ] }, { "modelId": "grok-4-3", "name": "Grok 4.3", "providerId": "xai", "releaseDate": null, "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 49.1, "Agents": null }, "coverage": 2, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 88.8, "percentile": 63.5, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" }, { "axisId": "frontiermath-1-3-epoch", "benchmark": "FrontierMath Tiers 1-3 v2 (Epoch AI run)", "category": "Reasoning", "rawScore": 42.8, "percentile": 34.7, "peerCount": 59, "sourceUrl": "https://epoch.ai/benchmarks/frontiermath", "measuredAt": "2026-09-17" } ] }, { "modelId": "glm-4-5", "name": "GLM-4.5", "providerId": "zai", "releaseDate": null, "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 76.3, "Agents": null }, "coverage": 1, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 93.3, "percentile": 76.3, "peerCount": 40, "sourceUrl": "https://matharena.ai/?comp=aime--aime_2025", "measuredAt": "2026-09-17" } ] }, { "modelId": "glm-4-5-air", "name": "GLM-4.5-Air", "providerId": "zai", "releaseDate": null, "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 33.8, "Agents": null }, "coverage": 1, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 83.3, "percentile": 33.8, "peerCount": 40, "sourceUrl": "https://matharena.ai/?comp=aime--aime_2025", "measuredAt": "2026-09-17" } ] }, { "modelId": "glm-4-6", "name": "GLM-4.6", "providerId": "zai", "releaseDate": null, "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 68.8, "Agents": null }, "coverage": 1, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "aime-2025", "benchmark": "AIME 2025", "category": "Reasoning", "rawScore": 91.7, "percentile": 68.8, "peerCount": 40, "sourceUrl": "https://matharena.ai/?comp=aime--aime_2025", "measuredAt": "2026-09-17" } ] }, { "modelId": "glm-4-7-flash", "name": "GLM-4.7-Flash", "providerId": "zai", "releaseDate": null, "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 11.3, "Agents": null }, "coverage": 1, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 60.5, "percentile": 11.3, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" } ] }, { "modelId": "qwen3-5-35b-a3b", "name": "Qwen3.5 35B-A3B", "providerId": "qwen", "releaseDate": null, "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": 43.7, "Agents": null }, "coverage": 1, "categoryCoverage": 1, "missingCategories": [ "Coding", "Agents" ], "components": [ { "axisId": "gpqa", "benchmark": "GPQA Diamond", "category": "Reasoning", "rawScore": 83.5, "percentile": 43.7, "peerCount": 111, "sourceUrl": "https://epoch.ai/benchmarks/gpqa-diamond", "measuredAt": "2026-09-17" } ] }, { "modelId": "glm-4-5-airx", "name": "GLM-4.5-AirX", "providerId": "zai", "releaseDate": null, "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "glm-4-5-flash", "name": "GLM-4.5-Flash", "providerId": "zai", "releaseDate": null, "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "glm-4-5-x", "name": "GLM-4.5-X", "providerId": "zai", "releaseDate": null, "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "glm-4-7-flashx", "name": "GLM-4.7-FlashX", "providerId": "zai", "releaseDate": null, "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "glm-5-turbo", "name": "GLM-5-Turbo", "providerId": "zai", "releaseDate": null, "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "mimo-v2-5", "name": "MiMo-V2.5", "providerId": "xiaomi", "releaseDate": null, "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "mimo-v2-5-pro", "name": "MiMo-V2.5-Pro", "providerId": "xiaomi", "releaseDate": null, "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "north-small-translate", "name": "North Small Translate", "providerId": "cohere", "releaseDate": null, "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "qwen3-5-122b-a10b", "name": "Qwen3.5 122B-A10B", "providerId": "qwen", "releaseDate": null, "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] }, { "modelId": "sonar-reasoning-pro", "name": "Sonar Reasoning Pro", "providerId": "perplexity", "releaseDate": null, "rank": null, "score": null, "categoryScores": { "Coding": null, "Reasoning": null, "Agents": null }, "coverage": 0, "categoryCoverage": 0, "missingCategories": [ "Coding", "Reasoning", "Agents" ], "components": [] } ] }, "meta": { "source": "https://llmmetric.com", "attribution": "Please cite llmmetric.com. Upstream records remain subject to their source terms.", "generatedAt": "2026-09-22T22:54:59.398Z" } }