Coding
NL2Repo leaderboard
As of 2026-11-21, the highest measured score on NL2Repo is 68.2% by DeepSeek V4.1 Flash 1017. 214 more models have estimated scores, calibrated from the benchmarks they were measured on.
Measured scores: benchlm.ai.
| # | Model | Score | Source |
|---|---|---|---|
| 1 | Gemini 4 Neon | 75.1% | estimated ± 6.5 pp, low confidence |
| 2 | GPT-6.2 Astra | 75.0% | estimated ± 6.5 pp, low confidence |
| 3 | Gemini 4 Carbon | 73.8% | estimated ± 6.5 pp, low confidence |
| 4 | GPT-6.2 Sol | 73.1% | estimated ± 6.5 pp, low confidence |
| 5 | Claude Mythos 5 | 71.7% | estimated ± 6.1 pp, low confidence |
| 6 | DeepSeek V4.1 Flash 1017 | 68.2% | measured |
| 7 | ERNIE 5.5 Flash Preview | 64.8% | measured |
| 8 | Muse Wonder Full | 64.7% | estimated ± 6.0 pp, low confidence |
| 9 | Muse Wonder | 64.6% | estimated ± 6.0 pp, low confidence |
| 10 | Gemini 4 Hydrogen | 64.6% | estimated ± 6.5 pp, low confidence |
| 11 | Gemini 4 Argon | 64.2% | estimated ± 6.0 pp, low confidence |
| 12 | Qwen4 Max | 63.3% | estimated ± 6.0 pp, medium confidence |
| 13 | Kimi K3.1 Pro | 63.2% | estimated ± 6.0 pp, medium confidence |
| 14 | Ember-1 | 63.0% | estimated ± 6.0 pp, medium confidence |
| 15 | Hy4 | 62.9% | estimated ± 6.0 pp, medium confidence |
| 16 | Claude Fable 5.2 | 62.7% | estimated ± 6.0 pp, medium confidence |
| 17 | Claude Opus 5.5 | 62.6% | estimated ± 6.0 pp, medium confidence |
| 18 | Pareto 26.9 | 62.5% | estimated ± 6.0 pp, medium confidence |
| 19 | MAI Code 2 | 62.2% | estimated ± 6.0 pp, medium confidence |
| 20 | Sakana Fugu-Ultra | 62.1% | estimated ± 2.8 pp, medium confidence |
| 21 | SWE-2 | 62.1% | estimated ± 6.0 pp, medium confidence |
| 22 | Particle-1 | 61.9% | estimated ± 6.0 pp, medium confidence |
| 23 | Inkling 2 Small | 61.7% | estimated ± 6.0 pp, medium confidence |
| 24 | GPT-6.1 Sol | 61.6% | estimated ± 6.0 pp, medium confidence |
| 25 | MiMo-V2.6-Pro | 61.6% | estimated ± 6.0 pp, medium confidence |
| 26 | DeepSeek V4 Pro | 61.5% | measured |
| 27 | Claude Sonnet 5.5 | 61.2% | estimated ± 6.0 pp, medium confidence |
| 28 | Grok 4.7 | 61.2% | estimated ± 6.0 pp, medium confidence |
| 29 | MiniMax M3.1 Flash | 61.2% | measured |
| 30 | Claude Opus 5 | 61.1% | estimated ± 4.7 pp, medium confidence |
| 31 | Claude Fable 5.1 | 61.0% | estimated ± 4.7 pp, medium confidence |
| 32 | Pareto 26.10 Preview | 60.8% | estimated ± 6.0 pp, medium confidence |
| 33 | LongCat-2.5 | 60.5% | measured |
| 34 | GPT-6 Sol | 60.3% | estimated ± 6.0 pp, medium confidence |
| 35 | GPT-5.6 Sol | 60.2% | estimated ± 4.7 pp, high confidence |
| 36 | MiMo-V2.6-Flash | 59.9% | estimated ± 6.0 pp, medium confidence |
| 37 | Step 5 Preview | 59.8% | estimated ± 6.0 pp, medium confidence |
| 38 | Grok 4.6 | 59.6% | estimated ± 4.7 pp, high confidence |
| 39 | Ornith-1.5-397B | 59.5% | measured |
| 40 | GPT-6 Luna | 59.4% | estimated ± 6.0 pp, medium confidence |
| 41 | GPT-5.6 Terra | 59.3% | estimated ± 4.7 pp, high confidence |
| 42 | Claude Fable 5 | 58.9% | estimated ± 4.7 pp, high confidence |
| 43 | Hy4 preview | 58.9% | measured |
| 44 | GPT-6.1 Terra | 58.7% | estimated ± 7.4 pp, medium confidence |
| 45 | Hy4 flash | 58.5% | estimated ± 6.0 pp, medium confidence |
| 46 | GLM-5.3 | 58.0% | measured |
| 47 | GPT-6 Astra | 57.2% | estimated ± 4.7 pp, medium confidence |
| 48 | Kimi K3 | 57.2% | estimated ± 4.7 pp, high confidence |
| 49 | Ultramarine 1 Flash | 57.2% | estimated ± 6.0 pp, medium confidence |
| 50 | GPT-5.6 Luna | 56.8% | estimated ± 4.7 pp, high confidence |
| 51 | SWE-2 Flash | 56.6% | estimated ± 6.0 pp, medium confidence |
| 52 | Muse Spark Open | 56.5% | estimated ± 6.0 pp, medium confidence |
| 53 | MAI Code 2 Flash | 56.5% | estimated ± 6.0 pp, medium confidence |
| 54 | Ling 3.1 Flash | 56.4% | estimated ± 6.0 pp, medium confidence |
| 55 | Muse Spark 1.3 | 56.3% | estimated ± 4.7 pp, medium confidence |
| 56 | GLM-5.3-Flash | 56.3% | measured |
| 57 | Claude Haiku 5.5 | 56.1% | estimated ± 6.5 pp, low confidence |
| 58 | Qwen3.8 Max | 55.9% | measured |
| 59 | DeepSeek V4 Flash 0731 | 54.2% | measured |
| 60 | Qwen3.8 Max Preview | 53.1% | estimated ± 4.7 pp, high confidence |
| 61 | Claude Opus 4.8 | 52.4% | estimated ± 4.7 pp, high confidence |
| 62 | Inkling 2 Mini | 51.5% | estimated ± 6.0 pp, medium confidence |
| 63 | Mercury 3 | 50.7% | estimated ± 6.0 pp, medium confidence |
| 64 | Grok 4.5 | 50.6% | estimated ± 4.7 pp, high confidence |
| 65 | Muse Spark 1.2 | 50.6% | estimated ± 4.7 pp, high confidence |
| 66 | Ripple-65B Preview | 49.8% | estimated ± 6.0 pp, medium confidence |
| 67 | dots3-note Preview | 49.8% | measured |
| 68 | Beam 1.0 501B-A23B | 49.4% | estimated ± 6.0 pp, medium confidence |
| 69 | Beam 1.0 501B-A23B FP8 | 49.3% | estimated ± 6.0 pp, medium confidence |
| 70 | Beam 1.0 501B-A23B NVFP4 | 49.1% | estimated ± 6.0 pp, medium confidence |
| 71 | GLM-5.2 | 48.9% | measured |
| 72 | Qwen3.8-Omni-Flash | 48.9% | measured |
| 73 | Sage-A1.5 | 48.4% | estimated ± 6.1 pp, medium confidence |
| 74 | Solar Open 2 | 48.3% | estimated ± 3.4 pp, high confidence |
| 75 | Ornith-1.0-397B | 48.2% | measured |
| 76 | Laguna S 2.1 | 48.2% | estimated ± 6.0 pp, medium confidence |
| 77 | Qwen3.8-Flash-Next | 48.1% | measured |
| 78 | GPT-5.5 | 47.4% | estimated ± 4.7 pp, high confidence |
| 79 | Qwen3.7 Max | 47.2% | measured |
| 80 | Inkling-Small | 47.1% | estimated ± 4.7 pp, high confidence |
| 81 | Seed 2.1 Pro | 47.0% | measured |
| 82 | Claude Opus 4.7 | 46.9% | estimated ± 4.7 pp, high confidence |
| 83 | Muse Spark 1.1 | 46.9% | estimated ± 4.7 pp, high confidence |
| 84 | Atria Dawn Preview | 46.4% | estimated ± 6.1 pp, medium confidence |
| 85 | Ornith-1.5-35B-A3B | 46.2% | measured |
| 86 | GPT-5.1-Codex | 46.1% | estimated ± 7.9 pp, low confidence |
| 87 | Gemini 3.7 Flash | 46.0% | estimated ± 4.7 pp, high confidence |
| 88 | o3 | 45.7% | estimated ± 7.9 pp, low confidence |
| 89 | GPT-5.1-Codex-Max | 45.7% | estimated ± 7.9 pp, low confidence |
| 90 | o4-mini | 45.7% | estimated ± 7.9 pp, low confidence |
| 91 | GLM-4.6 | 45.7% | estimated ± 7.9 pp, low confidence |
| 92 | Qwen3 Max | 45.7% | estimated ± 7.9 pp, low confidence |
| 93 | o3-mini | 45.7% | estimated ± 7.9 pp, low confidence |
| 94 | Claude 3.7 Sonnet Thinking | 45.7% | estimated ± 7.9 pp, low confidence |
| 95 | GLM-4.5 | 45.7% | estimated ± 7.9 pp, low confidence |
| 96 | Gemini 3.8 Flash | 45.5% | estimated ± 4.7 pp, high confidence |
| 97 | Claude Sonnet 5 | 45.2% | estimated ± 4.7 pp, high confidence |
| 98 | Composer 2.5 | 45.2% | estimated ± 4.7 pp, high confidence |
| 99 | Gemini 3.6 Flash | 45.2% | estimated ± 4.7 pp, high confidence |
| 100 | Gemini 3.1 Pro | 44.7% | estimated ± 4.7 pp, high confidence |
| 101 | Gemini 3.5 Flash | 44.7% | estimated ± 4.7 pp, high confidence |
| 102 | Quasar 438B | 44.6% | estimated ± 4.7 pp, high confidence |
| 103 | Ternary Bonsai 2 27B | 44.4% | estimated ± 3.4 pp, high confidence |
| 104 | Claude Opus 4.6 (Adaptive) | 44.3% | estimated ± 4.7 pp, high confidence |
| 105 | GPT-5.4 | 44.3% | estimated ± 4.7 pp, high confidence |
| 106 | Kimi K2.7 Code | 44.3% | estimated ± 4.7 pp, high confidence |
| 107 | Apodex 1.1 | 44.3% | estimated ± 4.7 pp, high confidence |
| 108 | Apodex 1.1 Mini | 44.3% | estimated ± 4.7 pp, high confidence |
| 109 | GPT-5.3 Codex | 44.2% | estimated ± 4.7 pp, high confidence |
| 110 | Inkling | 43.9% | estimated ± 4.7 pp, high confidence |
| 111 | Claude Sonnet 4.6 | 43.8% | estimated ± 4.7 pp, high confidence |
| 112 | Sakana Fugu | 43.8% | estimated ± 2.8 pp, medium confidence |
| 113 | Seed 2.1 Turbo | 43.7% | measured |
| 114 | Claude Opus 4.5 Thinking | 43.2% | estimated ± 4.7 pp, high confidence |
| 115 | Gemini 3 Pro | 43.2% | estimated ± 4.7 pp, high confidence |
| 116 | Claude Opus 4.5 | 43.2% | measured |
| 117 | Qwen 3.6 Max (preview) | 42.9% | measured |
| 118 | GPT-5.2 | 42.9% | estimated ± 4.7 pp, high confidence |
| 119 | Kimi K2.6 | 42.8% | estimated ± 2.8 pp, high confidence |
| 120 | GLM-5.1 | 42.7% | measured |
| 121 | Hy3 | 42.7% | estimated ± 4.7 pp, high confidence |
| 122 | Hy3 Preview | 42.7% | estimated ± 4.7 pp, high confidence |
| 123 | Nemotron 3 Ultra | 42.6% | estimated ± 3.4 pp, high confidence |
| 124 | Gemini 3.5 Flash-Lite | 42.5% | estimated ± 4.7 pp, high confidence |
| 125 | Gemini 3 Flash | 42.5% | estimated ± 4.7 pp, high confidence |
| 126 | BTL-4 | 42.4% | estimated ± 6.2 pp, medium confidence |
| 127 | Qwen3.8-27B | 42.3% | measured |
| 128 | Muse Spark | 42.1% | estimated ± 4.7 pp, high confidence |
| 129 | MiniMax M3 | 42.1% | measured |
| 130 | MiMo-V2.5-Pro | 41.9% | estimated ± 4.7 pp, high confidence |
| 131 | MiMo-V2-Pro | 41.6% | estimated ± 6.2 pp, medium confidence |
| 132 | GPT-5.4 mini | 41.5% | estimated ± 4.7 pp, high confidence |
| 133 | GPT-5.2-Codex | 41.2% | estimated ± 4.7 pp, high confidence |
| 134 | BTL-3 | 41.1% | estimated ± 3.4 pp, high confidence |
| 135 | Qwen3.7 Plus | 41.1% | measured |
| 136 | Grok 4.20 | 41.1% | estimated ± 4.7 pp, high confidence |
| 137 | Grok 4.3 | 40.7% | estimated ± 4.7 pp, high confidence |
| 138 | MiMo-V2.5 | 40.5% | estimated ± 4.7 pp, high confidence |
| 139 | Claude Sonnet 4.5 Thinking | 40.1% | estimated ± 4.7 pp, high confidence |
| 140 | GPT-5.1 | 40.1% | estimated ± 4.7 pp, high confidence |
| 141 | GPT-5.4 nano | 40.1% | estimated ± 4.7 pp, high confidence |
| 142 | Spark-X2.6-4B | 40.0% | estimated ± 6.1 pp, medium confidence |
| 143 | Claude Sonnet 4.5 | 39.9% | estimated ± 6.2 pp, medium confidence |
| 144 | MiniMax M2.7 | 39.8% | measured |
| 145 | GPT-5 | 39.7% | estimated ± 4.7 pp, high confidence |
| 146 | Ling 3.0 Flash VL | 39.5% | estimated ± 7.4 pp, medium confidence |
| 147 | Claude Haiku 4.5 | 38.9% | estimated ± 4.7 pp, medium confidence |
| 148 | Claude Opus 4.6 | 38.9% | estimated ± 6.1 pp, medium confidence |
| 149 | Mistral Medium 3.5 128B | 38.8% | estimated ± 4.7 pp, medium confidence |
| 150 | Ling 3.0 Flash | 38.5% | estimated ± 4.7 pp, medium confidence |
| 151 | Step 3.5 Flash | 38.4% | estimated ± 3.4 pp, high confidence |
| 152 | Qwen3.5 Flash | 38.2% | estimated ± 4.7 pp, medium confidence |
| 153 | Qwen3.6 Plus | 37.9% | measured |
| 154 | Gemini 3.1 Flash-Lite | 37.8% | estimated ± 4.7 pp, medium confidence |
| 155 | Grok 4 | 36.8% | estimated ± 4.7 pp, medium confidence |
| 156 | Laguna M.1 | 36.8% | estimated ± 4.7 pp, medium confidence |
| 157 | Laguna XS.2 | 36.4% | estimated ± 4.7 pp, medium confidence |
| 158 | Gemini 2.5 Pro | 36.3% | estimated ± 4.7 pp, medium confidence |
| 159 | Qwen3.6-27B | 36.2% | measured |
| 160 | GLM-4.7 | 36.1% | estimated ± 3.4 pp, high confidence |
| 161 | GLM-5 | 35.9% | measured |
| 162 | Grok 4 Fast (Reasoning) | 35.6% | estimated ± 4.7 pp, medium confidence |
| 163 | Grok 4.1 Fast (Reasoning) | 35.4% | estimated ± 4.7 pp, medium confidence |
| 164 | Mistral Large 3 | 35.4% | estimated ± 4.7 pp, medium confidence |
| 165 | MAI-Thinking-1 | 35.4% | estimated ± 2.8 pp, high confidence |
| 166 | MiMo-V2-Omni | 35.0% | estimated ± 6.2 pp, medium confidence |
| 167 | Muse Glimmer 30B | 34.8% | estimated ± 4.7 pp, high confidence |
| 168 | A.X K2 | 34.8% | estimated ± 3.4 pp, high confidence |
| 169 | Ornith-1.0-35B | 34.6% | measured |
| 170 | Claude 4.1 Opus | 34.5% | estimated ± 6.2 pp, medium confidence |
| 171 | Nemotron 3.5 Super VL 120B A12B | 34.2% | estimated ± 2.8 pp, high confidence |
| 172 | Nemotron 3.5 Super VL 120B A12B NVFP4 | 33.5% | estimated ± 2.8 pp, high confidence |
| 173 | Ornith-1.5-9B | 32.4% | measured |
| 174 | Qwen3.5 397B | 32.2% | measured |
| 175 | Kimi K2.5 | 32.0% | measured |
| 176 | Mellum2.1-12B-A2.5B-Thinking | 31.8% | estimated ± 3.4 pp, high confidence |
| 177 | Laguna XS 2.1 | 31.8% | estimated ± 6.1 pp, medium confidence |
| 178 | Nemotron 3.5 Bolt 74B-A9B | 31.6% | estimated ± 7.4 pp, medium confidence |
| 179 | Jais-3-MoVA-36B-A4B | 31.2% | estimated ± 6.2 pp, medium confidence |
| 180 | Claude 4 Sonnet | 31.0% | estimated ± 6.2 pp, medium confidence |
| 181 | MAI-Code-1.1-Flash | 30.8% | estimated ± 6.2 pp, medium confidence |
| 182 | Claude 4 Opus | 30.6% | estimated ± 6.2 pp, medium confidence |
| 183 | Jais-3-32B | 30.4% | estimated ± 6.2 pp, medium confidence |
| 184 | DeepSeek V3 0324 | 30.4% | estimated ± 7.4 pp, medium confidence |
| 185 | Qwen3.5-27B | 30.0% | estimated ± 3.4 pp, high confidence |
| 186 | North Mini Code | 30.0% | estimated ± 7.4 pp, medium confidence |
| 187 | MiMo-V2-Flash | 29.9% | estimated ± 3.4 pp, high confidence |
| 188 | Mercury 2.5 | 29.5% | estimated ± 7.4 pp, medium confidence |
| 189 | Flint | 29.5% | estimated ± 3.4 pp, medium confidence |
| 190 | Qwen3.6-35B-A3B | 29.4% | measured |
| 191 | Gemma 4 31B | 29.1% | estimated ± 3.4 pp, medium confidence |
| 192 | Qwen3.5-122B-A10B | 27.7% | estimated ± 3.4 pp, medium confidence |
| 193 | Grok Code Fast 1 | 27.5% | estimated ± 6.2 pp, medium confidence |
| 194 | o1 | 27.4% | estimated ± 4.7 pp, medium confidence |
| 195 | Step 3.7 Flash | 27.3% | estimated ± 4.7 pp, medium confidence |
| 196 | Ornith-1.0-9B | 27.2% | measured |
| 197 | Solar Pro 4 | 27.1% | estimated ± 6.2 pp, medium confidence |
| 198 | Jais-3-375B-A23B | 26.4% | estimated ± 6.2 pp, medium confidence |
| 199 | Jais-3-7B | 26.4% | estimated ± 6.2 pp, medium confidence |
| 200 | Gemma 4 26B A4B | 25.4% | estimated ± 3.4 pp, medium confidence |
| 201 | Qwen3-Coder 480B-A35B | 25.4% | estimated ± 6.2 pp, medium confidence |
| 202 | Granite 4.2 30B | 23.9% | estimated ± 3.4 pp, medium confidence |
| 203 | LongCat-Flash-Lite-Sparse | 23.3% | estimated ± 6.1 pp, low confidence |
| 204 | K-EXAONE 2.0 | 23.1% | estimated ± 6.2 pp, low confidence |
| 205 | o1-preview | 22.8% | estimated ± 4.7 pp, medium confidence |
| 206 | Qwen3.5-35B-A3B | 22.6% | estimated ± 3.4 pp, medium confidence |
| 207 | LFM3-1.7B | 21.4% | estimated ± 3.4 pp, medium confidence |
| 208 | K-Exaone | 21.3% | estimated ± 4.7 pp, medium confidence |
| 209 | Granite 4.2 8B | 21.2% | estimated ± 3.4 pp, medium confidence |
| 210 | Gemma 4 12B | 20.0% | estimated ± 3.4 pp, medium confidence |
| 211 | GPT-OSS 120B | 19.9% | estimated ± 4.7 pp, medium confidence |
| 212 | DeepSeek V3.1 | 19.6% | estimated ± 6.2 pp, low confidence |
| 213 | Kimi K2 | 19.3% | estimated ± 6.2 pp, low confidence |
| 214 | Mellum2-12B-A2.5B-Thinking | 18.1% | estimated ± 3.4 pp, medium confidence |
| 215 | Granite 4.2 3B | 17.9% | estimated ± 3.4 pp, medium confidence |
| 216 | Command A+ | 17.9% | estimated ± 4.7 pp, medium confidence |
| 217 | MiniCPM5-2B | 17.4% | estimated ± 3.4 pp, medium confidence |
| 218 | Nemotron 3.5 Lightning 30B A3B NVFP4 | 17.0% | estimated ± 4.7 pp, medium confidence |
| 219 | Mistral Small 4 | 16.9% | estimated ± 4.7 pp, medium confidence |
| 220 | Nemotron 3 Nano 30B | 16.8% | estimated ± 3.4 pp, medium confidence |
| 221 | Trinity-Large-Preview | 16.2% | estimated ± 4.7 pp, medium confidence |
| 222 | Trinity-Large-Thinking | 16.2% | estimated ± 4.7 pp, medium confidence |
| 223 | Ling 2.6 Flash | 15.8% | estimated ± 4.7 pp, medium confidence |
| 224 | Claude 3.7 Sonnet | 14.6% | estimated ± 6.2 pp, low confidence |
| 225 | Gemini 1.5 Pro | 14.5% | estimated ± 4.7 pp, medium confidence |
| 226 | DeepSeek V3 | 14.0% | estimated ± 4.7 pp, medium confidence |
| 227 | GPT-4 Turbo | 12.8% | estimated ± 4.7 pp, medium confidence |
| 228 | Nemotron 3 Super 100B | 12.5% | estimated ± 6.2 pp, low confidence |
| 229 | GPT-OSS 20B | 12.1% | estimated ± 4.7 pp, medium confidence |
| 230 | GPT-4.1 mini | 11.7% | estimated ± 4.7 pp, medium confidence |
| 231 | Claude 3 Opus | 11.2% | estimated ± 4.7 pp, medium confidence |
| 232 | GLM-4.7-Flash | 11.2% | estimated ± 6.2 pp, low confidence |
| 233 | Mistral Small 3.1 | 10.7% | estimated ± 7.4 pp, low confidence |
| 234 | LLaDA2.2-flash | 10.5% | estimated ± 6.1 pp, low confidence |
| 235 | DeepSeek-R1-0528 | 9.6% | estimated ± 6.2 pp, low confidence |
| 236 | Llama 4 Maverick | 8.6% | estimated ± 4.7 pp, medium confidence |
| 237 | MiniMax M1 80k | 8.3% | estimated ± 6.2 pp, low confidence |
| 238 | GPT-4.1 | 7.2% | estimated ± 6.2 pp, low confidence |
| 239 | Celeris-1 | 7.1% | estimated ± 4.7 pp, medium confidence |
| 240 | Solar Pro 3 | 6.7% | estimated ± 7.4 pp, low confidence |
| 241 | Nemotron 3 Nano Omni 30B A3B | 6.6% | estimated ± 4.7 pp, medium confidence |
| 242 | ZAYA1-74B-Preview | 6.2% | estimated ± 6.2 pp, low confidence |
| 243 | Ultravox v0.6 Llama 3.3 70B | 5.1% | estimated ± 4.7 pp, medium confidence |
| 244 | GPT-4o mini | 4.6% | estimated ± 4.7 pp, medium confidence |
| 245 | GPT-4.1 nano | 4.5% | estimated ± 4.7 pp, medium confidence |
| 246 | Ling 3.0 Tiny | 4.4% | estimated ± 7.4 pp, low confidence |
| 247 | Gemma 3 27B | 3.6% | estimated ± 4.7 pp, medium confidence |
| 248 | Gemma 4 E4B | 3.1% | estimated ± 4.7 pp, medium confidence |
| 249 | LFM2.5-2.6B | 0.0% | estimated ± 7.4 pp, low confidence |
| 250 | Llama 4 Scout | 0.0% | estimated ± 7.4 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
- cursorBench41
- 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
- ZeroBench w/ Python
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
- 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