Knowledge & reasoning
HealthBench Professional (raw, Artificial Analysis) leaderboard
As of 2026-11-18, the highest measured score on HealthBench Professional (raw, Artificial Analysis) is 71.4% by Claude Opus 5.5. 205 more models have estimated scores, calibrated from the benchmarks they were measured on.
Measured scores: benchlm.ai.
| # | Model | Score | Source |
|---|---|---|---|
| 1 | Claude Opus 5.5 | 71.4% | measured |
| 2 | GPT-6.2 Astra | 71.4% | estimated ± 1.1 pp, medium confidence |
| 3 | Claude Fable 5.2 | 71.1% | measured |
| 4 | Gemini 4 Carbon | 71.0% | estimated ± 1.1 pp, medium confidence |
| 5 | Claude Sonnet 5.5 | 70.8% | measured |
| 6 | Muse Wonder Full | 70.7% | estimated ± 0.8 pp, medium confidence |
| 7 | GPT-5.6 Sol | 70.6% | estimated ± 0.8 pp, medium confidence |
| 8 | GPT-6.2 Sol | 70.5% | estimated ± 1.1 pp, medium confidence |
| 9 | GPT-6.1 Sol | 70.4% | estimated ± 0.8 pp, medium confidence |
| 10 | GPT-6 Sol | 70.2% | estimated ± 0.8 pp, medium confidence |
| 11 | GPT-5.5 Pro | 70.1% | estimated ± 0.8 pp, medium confidence |
| 12 | Gemini 4 Hydrogen | 69.9% | estimated ± 1.1 pp, medium confidence |
| 13 | GPT-5.4 Pro | 69.9% | estimated ± 0.8 pp, medium confidence |
| 14 | GPT-5.6 Terra | 69.9% | estimated ± 0.8 pp, medium confidence |
| 15 | GPT-6 Astra | 69.7% | measured |
| 16 | Claude Opus 5 | 69.6% | estimated ± 0.8 pp, medium confidence |
| 17 | Claude Mythos 5 | 69.4% | estimated ± 2.6 pp, medium confidence |
| 18 | Claude Fable 5 | 69.4% | estimated ± 0.8 pp, medium confidence |
| 19 | Muse Wonder | 69.3% | estimated ± 0.8 pp, medium confidence |
| 20 | GPT-6.1 Terra | 69.2% | estimated ± 0.8 pp, medium confidence |
| 21 | Claude Fable 5.1 | 69.2% | measured |
| 22 | Gemini 4 Argon | 68.9% | estimated ± 0.8 pp, medium confidence |
| 23 | GPT-5.5 | 68.9% | estimated ± 0.8 pp, medium confidence |
| 24 | MiMo-V2.6-Pro | 68.7% | estimated ± 0.8 pp, medium confidence |
| 25 | Muse Spark 1.3 | 68.0% | estimated ± 0.8 pp, medium confidence |
| 26 | Kimi K3.1 Pro | 67.9% | estimated ± 0.8 pp, medium confidence |
| 27 | GPT-5.4 | 67.3% | estimated ± 0.8 pp, medium confidence |
| 28 | Kimi K3 | 67.3% | estimated ± 0.8 pp, medium confidence |
| 29 | Qwen4 Max | 67.2% | estimated ± 1.1 pp, medium confidence |
| 30 | Ling 3.0 Tiny | 67.1% | estimated ± 2.5 pp, medium confidence |
| 31 | Sakana Fugu-Ultra | 67.1% | estimated ± 2.6 pp, medium confidence |
| 32 | MiniMax M3.1 Flash | 66.8% | estimated ± 0.8 pp, medium confidence |
| 33 | Pareto 26.9 | 66.8% | estimated ± 2.6 pp, medium confidence |
| 34 | Sakana Fugu | 66.2% | estimated ± 2.6 pp, medium confidence |
| 35 | Claude Opus 4.8 | 66.0% | estimated ± 0.8 pp, medium confidence |
| 36 | GLM-5.2 | 66.0% | estimated ± 0.8 pp, medium confidence |
| 37 | Step 5 Preview | 66.0% | estimated ± 0.8 pp, medium confidence |
| 38 | ERNIE 5.5 Flash Preview | 65.9% | estimated ± 2.6 pp, medium confidence |
| 39 | GPT-5.6 Luna | 65.9% | estimated ± 0.8 pp, medium confidence |
| 40 | Sage-A1.5 | 65.5% | estimated ± 0.8 pp, medium confidence |
| 41 | Ornith-1.5-397B | 65.3% | estimated ± 2.6 pp, low confidence |
| 42 | GPT-6 Luna | 65.1% | estimated ± 0.8 pp, medium confidence |
| 43 | Qwen3.8 Max | 64.9% | estimated ± 2.6 pp, low confidence |
| 44 | GLM-5.3 | 64.9% | estimated ± 0.8 pp, medium confidence |
| 45 | Muse Ember | 64.9% | estimated ± 1.1 pp, low confidence |
| 46 | Claude Haiku 5.5 | 64.8% | measured |
| 47 | Gemini 3.8 Flash | 64.4% | estimated ± 0.8 pp, low confidence |
| 48 | DeepSeek V4 Pro | 64.2% | estimated ± 0.8 pp, low confidence |
| 49 | Ling 3.1 Flash | 64.2% | estimated ± 0.8 pp, low confidence |
| 50 | Gemini 3.1 Pro | 63.9% | estimated ± 0.8 pp, low confidence |
| 51 | Grok 4.7 | 63.9% | estimated ± 0.8 pp, low confidence |
| 52 | Muse Spark 1.2 | 63.9% | estimated ± 0.8 pp, low confidence |
| 53 | Qwen3.8 Max Preview | 63.9% | estimated ± 0.8 pp, low confidence |
| 54 | Beam 1.0 501B-A23B FP8 | 63.7% | estimated ± 0.8 pp, low confidence |
| 55 | Grok 4.6 | 63.5% | estimated ± 0.8 pp, low confidence |
| 56 | Claude Sonnet 5 | 63.3% | estimated ± 0.8 pp, low confidence |
| 57 | GPT-5.3 Codex | 63.3% | estimated ± 0.8 pp, low confidence |
| 58 | Hy4 preview | 63.3% | estimated ± 0.8 pp, low confidence |
| 59 | DeepSeek V4 Flash 0731 | 63.1% | estimated ± 0.8 pp, low confidence |
| 60 | Beam 1.0 501B-A23B | 62.8% | estimated ± 0.8 pp, low confidence |
| 61 | Inkling 2 Small | 62.6% | estimated ± 1.1 pp, low confidence |
| 62 | Grok 4.5 | 62.0% | estimated ± 0.8 pp, low confidence |
| 63 | Beam 1.0 501B-A23B NVFP4 | 62.0% | estimated ± 0.8 pp, low confidence |
| 64 | Muse Spark 1.1 | 61.7% | estimated ± 0.8 pp, low confidence |
| 65 | Muse Spark Open | 61.5% | estimated ± 0.8 pp, low confidence |
| 66 | MiMo-V2-Pro | 61.5% | estimated ± 1.1 pp, low confidence |
| 67 | GLM-5 | 61.1% | estimated ± 1.1 pp, low confidence |
| 68 | DeepSeek V4.1 Flash 1017 | 60.9% | estimated ± 0.8 pp, low confidence |
| 69 | Gemini 3.7 Flash | 60.9% | estimated ± 0.8 pp, low confidence |
| 70 | GLM-5-Turbo | 60.3% | estimated ± 1.1 pp, low confidence |
| 71 | Qwen3.7 Max | 59.9% | estimated ± 0.8 pp, low confidence |
| 72 | Gemini 3.5 Flash | 59.6% | estimated ± 0.8 pp, low confidence |
| 73 | Grok 4.20 | 59.3% | estimated ± 2.6 pp, low confidence |
| 74 | Claude Sonnet 4.6 | 59.1% | estimated ± 1.1 pp, low confidence |
| 75 | Claude Opus 4.6 (Adaptive) | 59.0% | estimated ± 0.8 pp, low confidence |
| 76 | MiMo-V2-Omni | 58.6% | estimated ± 1.1 pp, low confidence |
| 77 | Claude Opus 4.5 | 58.4% | estimated ± 1.1 pp, low confidence |
| 78 | GLM-5V-Turbo | 58.3% | estimated ± 1.1 pp, low confidence |
| 79 | Claude Opus 4.7 | 58.2% | estimated ± 0.8 pp, low confidence |
| 80 | MiMo-V2.6-Flash | 58.2% | estimated ± 0.8 pp, low confidence |
| 81 | Qwen3.5-27B | 57.8% | estimated ± 1.1 pp, low confidence |
| 82 | MiniMax M2.7 | 57.7% | estimated ± 1.1 pp, low confidence |
| 83 | GPT-5.2 | 57.6% | estimated ± 0.8 pp, low confidence |
| 84 | Solar Open 2 | 57.6% | estimated ± 2.6 pp, low confidence |
| 85 | Command A+ | 57.5% | estimated ± 1.1 pp, low confidence |
| 86 | Grok 4 | 57.5% | estimated ± 1.1 pp, low confidence |
| 87 | GLM-4.7 | 57.3% | estimated ± 1.1 pp, low confidence |
| 88 | Muse Spark | 57.2% | estimated ± 0.8 pp, low confidence |
| 89 | o3-pro | 57.0% | estimated ± 1.1 pp, low confidence |
| 90 | Qwen3.8-Flash-Next | 56.9% | estimated ± 0.8 pp, low confidence |
| 91 | Qwen3.6-27B | 56.6% | estimated ± 1.1 pp, low confidence |
| 92 | Nemotron 3.5 Super VL 120B A12B | 56.5% | estimated ± 0.8 pp, low confidence |
| 93 | Gemini 3.6 Flash | 56.1% | estimated ± 0.8 pp, low confidence |
| 94 | Nemotron 3.5 Super VL 120B A12B NVFP4 | 56.1% | estimated ± 0.8 pp, low confidence |
| 95 | Claude 4 Opus Thinking | 56.0% | estimated ± 1.1 pp, low confidence |
| 96 | Ling 3.0 Flash | 55.5% | estimated ± 1.1 pp, low confidence |
| 97 | Ornith-1.5-35B-A3B | 55.3% | estimated ± 2.6 pp, low confidence |
| 98 | GPT-5.4 mini | 55.1% | estimated ± 0.8 pp, low confidence |
| 99 | Kimi K2.7 Code | 55.1% | estimated ± 0.8 pp, low confidence |
| 100 | Mistral Large 4 | 55.1% | estimated ± 0.8 pp, low confidence |
| 101 | Step 3.7 Flash | 54.9% | estimated ± 1.1 pp, low confidence |
| 102 | Qwen3.5-35B-A3B | 54.8% | estimated ± 1.1 pp, low confidence |
| 103 | Quasar 438B | 54.0% | estimated ± 0.8 pp, low confidence |
| 104 | Claude 4.1 Opus | 54.0% | estimated ± 1.1 pp, low confidence |
| 105 | GPT-5.4 nano | 53.8% | estimated ± 0.8 pp, low confidence |
| 106 | Gemini 3 Pro | 53.4% | estimated ± 0.8 pp, low confidence |
| 107 | Qwen3.7 Plus | 53.4% | estimated ± 0.8 pp, low confidence |
| 108 | Gemini 3 Flash | 53.4% | estimated ± 1.1 pp, low confidence |
| 109 | Claude 3.7 Sonnet Thinking | 53.1% | estimated ± 1.1 pp, low confidence |
| 110 | A.X K2 | 53.0% | estimated ± 0.8 pp, low confidence |
| 111 | GPT-5.2-Codex | 52.6% | estimated ± 0.8 pp, low confidence |
| 112 | o4-mini | 51.9% | estimated ± 1.1 pp, low confidence |
| 113 | Claude 4 Opus | 51.9% | estimated ± 1.1 pp, low confidence |
| 114 | Claude 4 Sonnet | 51.9% | estimated ± 1.1 pp, low confidence |
| 115 | Inkling-Small | 51.8% | estimated ± 0.8 pp, low confidence |
| 116 | MiMo-V2-Flash | 51.2% | estimated ± 1.1 pp, low confidence |
| 117 | DeepSeek V3.2 | 51.2% | estimated ± 1.1 pp, low confidence |
| 118 | Grok 4.3 | 51.1% | estimated ± 0.8 pp, low confidence |
| 119 | Kimi K2.6 | 51.1% | estimated ± 0.8 pp, low confidence |
| 120 | Qwen3 Max | 50.7% | estimated ± 1.1 pp, low confidence |
| 121 | Ornith-1.5-9B | 50.4% | estimated ± 2.6 pp, low confidence |
| 122 | Claude 3.7 Sonnet | 50.3% | estimated ± 1.1 pp, low confidence |
| 123 | o1 | 50.2% | estimated ± 1.1 pp, low confidence |
| 124 | GLM-4.6 | 49.8% | estimated ± 1.1 pp, low confidence |
| 125 | K-Exaone | 49.1% | estimated ± 1.1 pp, low confidence |
| 126 | Grok Code Fast 1 | 48.6% | estimated ± 1.1 pp, low confidence |
| 127 | Ling 2.6 Flash | 48.6% | estimated ± 1.1 pp, low confidence |
| 128 | DeepSeek V3.1 | 48.1% | estimated ± 1.1 pp, low confidence |
| 129 | DeepSeek-R1 | 47.2% | estimated ± 1.1 pp, low confidence |
| 130 | Kimi K2 | 46.6% | estimated ± 1.1 pp, low confidence |
| 131 | GPT-4.1 | 46.5% | estimated ± 1.1 pp, low confidence |
| 132 | Solar Pro 4 | 46.5% | estimated ± 0.8 pp, low confidence |
| 133 | o3-mini | 46.2% | estimated ± 1.1 pp, low confidence |
| 134 | o1-pro | 46.1% | estimated ± 1.1 pp, low confidence |
| 135 | Gemini 3.5 Flash-Lite | 44.8% | estimated ± 1.1 pp, low confidence |
| 136 | GPT-5 | 44.5% | estimated ± 0.8 pp, low confidence |
| 137 | GPT-5.1-Codex | 44.5% | estimated ± 0.8 pp, low confidence |
| 138 | GPT-5.1-Codex-Max | 44.5% | estimated ± 0.8 pp, low confidence |
| 139 | o1-preview | 44.3% | estimated ± 1.1 pp, low confidence |
| 140 | Grok 4.1 Fast | 44.1% | estimated ± 1.1 pp, low confidence |
| 141 | GLM-4.5-Air | 43.8% | estimated ± 1.1 pp, low confidence |
| 142 | Inkling | 43.4% | estimated ± 0.8 pp, low confidence |
| 143 | Qwen3.8-27B | 43.4% | estimated ± 0.8 pp, low confidence |
| 144 | GPT-4.1 mini | 42.0% | estimated ± 1.1 pp, low confidence |
| 145 | GPT-5.1 | 41.4% | estimated ± 0.8 pp, low confidence |
| 146 | Hy3 | 41.4% | estimated ± 0.8 pp, low confidence |
| 147 | Hy3 Preview | 41.4% | estimated ± 0.8 pp, low confidence |
| 148 | Gemini 2.5 Flash | 41.3% | estimated ± 1.1 pp, low confidence |
| 149 | o1-mini | 41.2% | estimated ± 1.1 pp, low confidence |
| 150 | DeepSeek V3 0324 | 41.1% | estimated ± 1.1 pp, low confidence |
| 151 | GPT-4.5 | 40.8% | estimated ± 1.1 pp, low confidence |
| 152 | QwQ-32B | 40.5% | estimated ± 1.1 pp, low confidence |
| 153 | Apodex 1.1 | 40.2% | estimated ± 0.8 pp, low confidence |
| 154 | Apodex 1.1 Mini | 40.2% | estimated ± 0.8 pp, low confidence |
| 155 | Claude Opus 4.5 Thinking | 40.2% | estimated ± 0.8 pp, low confidence |
| 156 | GLM-5.1 | 40.2% | estimated ± 0.8 pp, low confidence |
| 157 | Mistral Large 3 | 40.1% | estimated ± 1.1 pp, low confidence |
| 158 | Magistral Medium | 39.6% | estimated ± 1.1 pp, low confidence |
| 159 | Mistral Medium 3 | 39.6% | estimated ± 1.1 pp, low confidence |
| 160 | Claude 3.5 Haiku | 39.3% | estimated ± 1.1 pp, low confidence |
| 161 | Gemini 2.0 Flash | 39.3% | estimated ± 1.1 pp, low confidence |
| 162 | Claude 3 Opus | 38.8% | estimated ± 1.1 pp, low confidence |
| 163 | Nemotron 3.5 Bolt 74B-A9B | 38.8% | estimated ± 0.8 pp, low confidence |
| 164 | DeepSeek V3 | 38.3% | estimated ± 1.1 pp, low confidence |
| 165 | GPT-4o | 38.2% | estimated ± 1.1 pp, low confidence |
| 166 | MiMo-V2.5-Pro | 37.4% | estimated ± 0.8 pp, low confidence |
| 167 | MiniMax M3 | 35.8% | estimated ± 0.8 pp, low confidence |
| 168 | Qwen 3.6 Max (preview) | 35.8% | estimated ± 0.8 pp, low confidence |
| 169 | Mistral Medium 3.5 128B | 34.4% | estimated ± 2.5 pp, low confidence |
| 170 | Kimi K2.5 | 32.3% | estimated ± 0.8 pp, low confidence |
| 171 | Nemotron 3 Ultra | 32.3% | estimated ± 0.8 pp, low confidence |
| 172 | Grok 4.1 Fast (Reasoning) | 31.0% | estimated ± 0.8 pp, low confidence |
| 173 | Grok 4 Fast (Reasoning) | 31.0% | estimated ± 0.8 pp, low confidence |
| 174 | Qwen3.6 Plus | 31.0% | estimated ± 0.8 pp, low confidence |
| 175 | Qwen3.5 397B | 30.6% | estimated ± 2.1 pp, low confidence |
| 176 | Claude Opus 4.6 | 30.4% | estimated ± 0.8 pp, low confidence |
| 177 | Mercury 2.5 | 29.7% | estimated ± 2.5 pp, low confidence |
| 178 | Gemini 2.5 Pro | 29.0% | estimated ± 0.8 pp, low confidence |
| 179 | Muse Glimmer 30B | 29.0% | estimated ± 0.8 pp, low confidence |
| 180 | Nemotron 3.5 Bolt 74B-A9B BF16 | 28.8% | estimated ± 2.5 pp, low confidence |
| 181 | Mistral Small 4 | 26.9% | estimated ± 2.5 pp, low confidence |
| 182 | Mistral Large 2 | 20.1% | estimated ± 2.5 pp, low confidence |
| 183 | Qwen3.5-122B-A10B | 17.4% | estimated ± 2.1 pp, low confidence |
| 184 | Solar Pro 3 | 14.5% | estimated ± 2.5 pp, low confidence |
| 185 | North Mini Code | 14.1% | estimated ± 2.5 pp, low confidence |
| 186 | Gemma 4 31B | 12.7% | estimated ± 2.1 pp, low confidence |
| 187 | Magistral Small | 12.7% | estimated ± 2.5 pp, low confidence |
| 188 | Claude 3 Haiku | 11.3% | estimated ± 2.5 pp, low confidence |
| 189 | Qwen3.6-35B-A3B | 9.3% | estimated ± 2.1 pp, low confidence |
| 190 | Nova Pro | 9.2% | estimated ± 2.5 pp, low confidence |
| 191 | Ling 3.0 Flash VL | 8.9% | estimated ± 2.1 pp, low confidence |
| 192 | Solar Pro 2 | 7.1% | estimated ± 2.5 pp, low confidence |
| 193 | Gemini 2.5 Flash-Lite | 5.6% | estimated ± 2.5 pp, low confidence |
| 194 | o3 | 5.5% | estimated ± 2.1 pp, low confidence |
| 195 | Mistral Small 3.1 | 5.0% | estimated ± 2.5 pp, low confidence |
| 196 | GPT-OSS 120B | 4.7% | estimated ± 2.1 pp, low confidence |
| 197 | Gemma 4 26B A4B | 4.3% | estimated ± 2.1 pp, low confidence |
| 198 | Devstral Small | 4.3% | estimated ± 2.5 pp, low confidence |
| 199 | Nemotron 3.5 Lightning 30B A3B NVFP4 | 4.3% | estimated ± 2.5 pp, low confidence |
| 200 | K-EXAONE 2.0 | 3.5% | estimated ± 2.1 pp, low confidence |
| 201 | GPT-4.1 nano | 2.9% | estimated ± 2.5 pp, low confidence |
| 202 | Gemma 3 27B | 2.1% | estimated ± 2.5 pp, low confidence |
| 203 | Trinity-Large-Preview | 1.4% | estimated ± 2.1 pp, low confidence |
| 204 | Trinity-Large-Thinking | 1.4% | estimated ± 2.1 pp, low confidence |
| 205 | Gemma 4 12B | 1.3% | estimated ± 2.1 pp, low confidence |
| 206 | Granite 4.2 8B | 1.2% | estimated ± 2.5 pp, low confidence |
| 207 | Granite 4.2 30B | 0.9% | estimated ± 2.5 pp, low confidence |
| 208 | Gemma 4 E4B | 0.4% | estimated ± 2.5 pp, low confidence |
| 209 | LFM2.5-8B-A1B | 0.3% | estimated ± 2.5 pp, low confidence |
| 210 | Granite 4.2 3B | 0.3% | estimated ± 2.5 pp, low confidence |
| 211 | LFM3-1.7B | 0.1% | estimated ± 2.5 pp, low confidence |
All leaderboards
Agentic · terminal
- AA Terminal-Bench 2.1
- AA Terminal-Bench 4.0
- Terminal-Bench 4.0
- Terminal-Bench 2.0
- Terminal-Bench 2.1
- Terminal-Bench 3.0
- Terminal-Bench Hard
- Terminal-Bench-Science 0.1
- Terminal-Bench 4.1
- Terminal-Bench 2.1 (Vals AI)
Agentic · tools
- AA Agentic Index
- AA-AnalystAgent
- AA AutomationBench
- CWE-Bench-AA
- AA Cyber Index
- DeepsecBench-AA
- AA EnterpriseOps-Gym
- GDP.pdf
- AA Harvey LAB v1.0
- Harvey LAB-AA v1.1
- AA ITBench
- AA Tau3 Banking
- Agents' Last Exam
- APEX-Agents
- APEX-Agents-AA
- ApprenticeBench
- AutomationBench
- BFCL v4
- BrowseComp
- Claw-Eval
- CWE-bench v1
- CWE-bench
- CyberGym
- DeepPlanning
- DeepSearchQA
- DRACO
- ExploitGym
- Finance Agent v2
- GDPval-AA
- Gert Labs
- Hermes Bench
- Hermes Index
- HLE w/ tools
- JobBench
- LifeVal
- MazeBench
- MCP Atlas
- MCP-Tasks
- MKA
- OSWorld 2.1
- OSWorld 2.1 Offline
- OSWorld-Verified
- OSWorld 2.0
- PinchBench
- QwenClawBench
- ResearchClawBench
- SkillsBench
- τ²-bench results
- τ³-bench results
- Toolathlon
- Toolathlon-Verified
- VITA-Bench
- WideResearch
Coding
- AA Coding Index
- AA LiveCodeBench
- AA-SciCode
- Bug Hunt Bench
- CursorBench 3.1
- CursorBench 3.2
- CursorBench 4.0
- DeepSWE
- FrontierCode 1.1 Main
- FrontierCode 2.0
- FrontierCode 1.1 Extended
- FrontierSWE v2
- LiveCodeBench
- LiveCodeBench Pro
- LiveCodeBench v6
- NL2Repo
- OpenHarmony Bench v1.2
- OpenHarmony Bench v1.0
- PostTrainBench
- PostTrainBench v1
- PostTrainBench v1.1
- ProgramBench
- React Native Evals
- ScarfBench
- SciCode
- SWE-Atlas Codebase QnA
- SWE-bench Pro
- SWE-bench Verified
- SWE-Rebench
- SWE-Marathon
- Vals LiveCodeBench
- Vals SWE-bench
- Vibe Code Bench
- VulcanBench Frontier v4
- VulcanBench v3
Math
- AIME 2024
- AIME 2025
- AIME26
- FrontierMath v2 (Tier 4)
- FrontierMath v2 (Tiers 1-3)
- HMMT Feb 2025
- HMMT Feb 2026
- HMMT Nov 2025
- IMOAnswerBench
- MATH-500
- MMAnswerBench
Knowledge & reasoning
- AA-GPQA Diamond
- HealthBench Professional (length-adjusted, Artificial Analysis)
- HealthBench Professional (raw, Artificial Analysis)
- AA-HLE
- AA MMLU-Pro
- ARC-AGI-2
- ARC-AGI-3 Provider Adapter
- ARC-AGI-1
- ARC-AGI-3
- Artificial Analysis Intelligence Index
- BBH
- BioBasicsBench
- BioMysteryBench (human-difficult)
- BioMysteryBench (human-solvable)
- C-Eval
- CritPt
- GPQA
- GPQA Diamond
- HealthBench (raw)
- HealthBench Hard
- HealthBench (length-adjusted)
- HealthBench Professional
- HealthBench Professional (raw)
- HLE
- HLE w/o tools
- HLE-Verified
- JevBench 1.4
- LABBench2
- MedXpertQA (Text)
- MMLU
- MMLU-Pro
- MMLU-Pro (Arcee)
- MMLU-Redux
- MMMLU
- AA-Omniscience Accuracy
- SimpleQA
- STEMBench
- SuperGPQA
- Vals GPQA Diamond
- Vals MMLU-Pro
Instruction following
Multilingual
Vision & documents
- AA-MMMU-Pro
- AI2D_TEST
- BabyVision
- BabyVision w/ Python
- BenchCAD Vision2Code (no tools)
- BenchCAD Vision2Code (tools)
- CC-OCR
- Chartography (no tools)
- Chartography (tools)
- CharXiv
- CharXiv w/o tools
- CountBench
- DynaMath
- ERQA
- LVBench
- MathVision
- MathVision w/ Python
- MedXpertQA (MM)
- MLVU (M-Avg)
- MMMU
- MMMU-Pro
- MMMU-Pro w/ Python
- MMVU
- OCRBench V2
- OfficeQA
- OfficeQA Pro
- OmniDocBench 1.5
- ParseBench-1
- ParseBench-2
- PerceptionBench
- RealWorldQA
- RefCOCO (avg)
- ScreenSpot Pro
- SimpleVQA
- SWE-bench Multimodal
- Video-MME (w/o subtitle)
- Video-MME (with subtitle)
- VideoMMMU
- V*
- We-Math
- ZeroBench
Long context
General
- HLE-MultiChoice (Decision Index)
- Artificial Analysis Decision Index v1.0
- Artificial Analysis Decision Index
- When2Call (Decision Index)
- WildGuardTest (Decision Index)
- BabaIsBench
- BinaryAudit
- BlenderWorld (text-only)
- Chess Puzzles
- CompileBench
- Cookie-Bench
- DAYJOB
- DAYJOB Healthcare
- Decision Index Full Score
- EBR-bench
- Epoch Automation Reports
- ExploitBench
- ExploitBench Internal Port
- FACTS Multimodal
- FailureSensorIQ
- FrontierPhysics
- Furniture Assembly
- Decision Index Arts
- Decision Index Knowledge
- Decision Index Language
- Decision Index Retrieval
- Decision Index Tools
- MaxxBench
- MirrorCode
- Mystery Game Puzzles
- NYT Connections
- OTelBench
- SEC-Bench Pro
- SkillsBench official leaderboard (BenchFlow)
- SpireClimb
- StealthMark