| 1 | Claude Opus 5.5 | Anthropic | Proprietary | 46.4% | |
| 2 | Claude Fable 5.1 | Anthropic | Proprietary | 43.5% | |
| 3 | GPT-6 Astra | OpenAI | Proprietary | 43.4% | |
| 4 | Claude Fable 5 | Anthropic · Closed | 43.3% | | |
| 5 | Claude Opus 5 | Anthropic · Closed | 37.1% | | |
| 6 | Grok 4.7 | xAI · Closed | 32.0% | | |
| 7 | Gemini 3.1 Pro | Google · Closed | 31.9% | | |
| 8 | Grok 4.6 | xAI · Closed | 30.5% | | |
| 9 | Gemini 3.8 Flash | Google · Closed | 29.6% | | |
| 10 | Claude Opus 4.8 | Anthropic · Closed | 28.8% | | |
| 11 | Muse Spark 1.1 | Meta · Closed | 28.1% | | |
| 12 | Claude Opus 4.7 (Adaptive) | Anthropic · Closed | 27.3% | | |
| 13 | Muse Spark 1.2 | Meta · Closed | 27.2% | | |
| 14 | GPT-6 Sol | OpenAI · Closed | 27.1% | | |
| 15 | Gemini 3.7 Flash | Google · Closed | 26.5% | | |
| 16 | Grok 4.5 | xAI · Closed | 25.3% | | |
| 17 | Muse Spark 1.3 | Meta · Closed | 25.0% | | |
| 18 | Gemini 3.6 Flash | Google · Closed | 22.1% | | |
| 19 | GPT-5.6 Sol | OpenAI · Closed | 22.0% | | |
| 20 | Gemini 3.5 Flash | Google · Closed | 21.2% | | |
| 21 | GPT-5.5 | OpenAI · Closed | 20.5% | | |
| 22 | Kimi K3 | Moonshot AI · Closed | 19.7% | | |
| 23 | Grok 4.3 | xAI · Closed | 18.0% | | |
| 24 | Claude Sonnet 5 | Anthropic · Closed | 16.5% | | |
| 25 | Step 5 Preview | StepFun · Closed | 16.4% | | |
| 26 | Gemini 3 Pro | Google · Closed | 15.3% | | |
| 27 | Claude Opus 4.7 | Anthropic · Closed | 14.8% | | |
| 28 | GLM-5.3 | Z.AI · Open weight | 14.3% | | |
| 29 | Claude Opus 4.5 Thinking | Anthropic · Closed | 14.0% | | |
| 30 | Claude Opus 4.6 (Adaptive) | Anthropic · Closed | 13.7% | | |
| 31 | Qwen3.7 Max | Alibaba · Closed | 13.5% | | |
| 32 | Qwen3.8 Max Preview | Alibaba · Closed | 12.0% | | |
| 33 | GPT-5.3 Codex | OpenAI · Closed | 10.9% | | |
| 34 | GPT-5.3-Codex-Spark | OpenAI · Closed | 10.9% | | |
| 35 | Qwen 3.6 Max (preview) | Alibaba · Closed | 9.2% | | |
| 36 | MiMo-V2.6-Pro | Xiaomi · Open weight | 8.4% | | |
| 37 | Muse Spark | Meta · Closed | 7.2% | | |
| 38 | GPT-5.4 | OpenAI · Closed | 5.8% | | |
| 39 | GPT-5.1 | OpenAI · Closed | 5.4% | | |
| 40 | Kimi K2.6 | Moonshot AI · Open weight | 5.3% | | |
| 41 | Gemini 3.5 Flash-Lite | Google · Closed | 5.2% | | |
| 42 | MiMo-V2-Pro | Xiaomi · Closed | 4.6% | | |
| 43 | GLM-5.2 | Z.AI · Open weight | 4.4% | | |
| 44 | MiMo-V2.5-Pro | Xiaomi · Closed | 3.3% | | |
| 45 | Claude Opus 4.6 | Anthropic · Closed | 2.4% | | |
| 46 | Grok 4 | xAI · Closed | 2.1% | | |
| 47 | Inkling | Thinking Machines Lab · Open weight | 2.0% | | |
| 48 | MiniMax M3 | MiniMax · Open weight | 1.4% | | |
| 49 | Qwen3.7 Plus | Alibaba · Closed | 1.1% | | |
| 50 | GLM-5.1 | Z.AI · Open weight | 0.9% | | |
| 51 | Qwen3.6 Plus | Alibaba · Closed | 0.9% | | |
| 52 | DeepSeek V4 Pro 0813 | DeepSeek · Closed | 0.8% | | |
| 53 | MiniMax M2.7 | MiniMax · Open weight | 0.8% | | |
| 54 | GPT-6 Luna | OpenAI · Closed | 0.7% | | |
| 55 | GLM-5 | Z.AI · Open weight | 0.3% | | |
| 56 | GPT-5.6 Terra | OpenAI · Closed | 0.1% | | |
| 57 | Nemotron 3 Ultra | NVIDIA · Open weight | -0.4% | | |
| 58 | Solar Pro 4 | Upstage · Closed | -0.8% | | |
| 59 | GPT-5.2 | OpenAI · Closed | -0.9% | | |
| 60 | GPT-5.2-Codex | OpenAI · Closed | -2.2% | | |
| 61 | Quasar 438B | Multiverse Computing · Closed | -2.6% | | |
| 62 | Claude Sonnet 4.6 | Anthropic · Closed | -3.5% | | |
| 63 | Command A+ | Cohere · Open weight | -4.0% | | |
| 64 | Claude Opus 4.5 | Anthropic · Closed | -4.1% | | |
| 65 | Gemini 3 Flash | Google · Closed | -4.3% | | |
| 66 | Ling 3.0 Flash VL | InclusionAI · Open weight | -4.5% | | |
| 67 | DeepSeek V4.1 Flash | DeepSeek · Open weight | -5.3% | | |
| 68 | GPT-5.1-Codex | OpenAI · Closed | -6.5% | | |
| 69 | GPT-5.1-Codex-Max | OpenAI · Closed | -6.5% | | |
| 70 | K-EXAONE 2.0 | LG AI Research · Open weight | -6.6% | | |
| 71 | Kimi K2.5 | Moonshot AI · Open weight | -7.3% | | |
| 72 | Kimi K2.5 (Reasoning) | Moonshot AI · Closed | -7.3% | | |
| 73 | A.X K2 | SK Telecom · Open weight | -8.2% | | |
| 74 | GPT-5 (high) | OpenAI · Closed | -8.7% | | |
| 75 | Inkling-Small | Thinking Machines Lab · Open weight | -8.9% | | |
| 76 | Claude 4 Sonnet | Anthropic · Closed | -9.0% | | |
| 77 | Qwen3.8-Flash-Next | Alibaba · Open weight | -9.7% | | |
| 78 | Qwen3.8-27B | Alibaba · Open weight | -10.0% | | |
| 79 | Kimi K2.7 Code | Moonshot AI · Open weight | -10.2% | | |
| 80 | GPT-5.6 Luna | OpenAI · Closed | -10.3% | | |
| 81 | GPT-4o | OpenAI · Closed | -10.5% | | |
| 83 | GPT-5 (medium) | OpenAI · Closed | -10.9% | | |
| 84 | o1 | OpenAI · Closed | -11.0% | | |
| 85 | MiniCPM5-2B | OpenBMB · Open weight | -11.6% | | |
| 86 | Granite 4.2 30B | IBM · Open weight | -12.9% | | |
| 87 | DeepSeek V4 Flash 0731 | DeepSeek · Closed | -14.3% | | |
| 88 | Granite 4.2 3B | IBM · Open weight | -14.7% | | |
| 89 | o3 | OpenAI · Closed | -15.5% | | |
| 90 | Gemini 2.5 Pro | Google · Closed | -16.3% | | |
| 91 | GLM-5-Turbo | Z.AI · Closed | -16.4% | | |
| 92 | Llama 3.1 405B | Meta · Open weight | -17.1% | | |
| 93 | Granite 4.2 8B | IBM · Open weight | -17.2% | | |
| 94 | GPT-5 mini | OpenAI · Closed | -17.3% | | |
| 95 | Nemotron 3.5 Lightning 30B A3B NVFP4 | NVIDIA · Open weight | -17.7% | | |
| 96 | Ling 3.0 Flash | InclusionAI · Open weight | -17.9% | | |
| 97 | Ling 3.0 Flash FP8 | InclusionAI · Open weight | -17.9% | | |
| 98 | Hy3 | Tencent · Open weight | -18.5% | | |
| 99 | Hy3 Preview | Tencent · Open weight | -18.5% | | |
| 100 | GPT-5.4 mini | OpenAI · Closed | -18.9% | | |
| 101 | GLM-5V-Turbo | Z.AI · Closed | -19.3% | | |
| 102 | Ling 3.0 Tiny | InclusionAI · Open weight | -19.3% | | |
| 103 | Gemma 4 E4B | Google · Open weight | -19.7% | | |
| 104 | Qwen3.6-27B | Alibaba · Open weight | -20.0% | | |
| 105 | MiMo-V2-Omni | Xiaomi · Closed | -20.1% | | |
| 106 | Apodex 1.1 | Apodex · Closed | -21.9% | | |
| 107 | Apodex 1.1 Mini | Apodex · Open weight | -21.9% | | |
| 108 | Qwen3.6-35B-A3B | Alibaba · Open weight | -22.2% | | |
| 109 | Gemma 4 E2B | Google · Open weight | -23.6% | | |
| 110 | DeepSeek-R1 | DeepSeek · Open weight | -27.4% | | |
| 111 | Kimi K2 | Moonshot AI · Closed | -28.3% | | |
| 112 | GPT-5 nano | OpenAI · Closed | -28.7% | | |
| 113 | GPT-5.4 nano | OpenAI · Closed | -29.5% | | |
| 114 | LFM2.5-2.6B | LiquidAI · Open weight | -29.5% | | |
| 115 | DeepSeek V3.1 (Reasoning) | DeepSeek · Open weight | -29.6% | | |
| 116 | Grok 4.1 Fast (Reasoning) | xAI · Closed | -29.9% | | |
| 117 | Grok 4 Fast (Reasoning) | xAI · Closed | -29.9% | | |
| 118 | Mistral Small 4 | Mistral · Open weight | -30.4% | | |
| 119 | Mistral Small 4 (Reasoning) | Mistral · Open weight | -30.4% | | |
| 120 | Mistral Medium 3 | Mistral · Closed | -31.4% | | |
| 121 | GLM-4.6 | Z.AI · Open weight | -31.7% | | |
| 122 | Muse Glimmer 30B | Meta · Open weight | -32.8% | | |
| 123 | LFM2.5-8B-A1B | LiquidAI · Open weight | -33.3% | | |
| 124 | Mistral Large 2 | Mistral · Closed | -34.4% | | |
| 125 | GLM-4.7 | Z.AI · Open weight | -36.4% | | |
| 126 | Mistral Medium 3.5 128B | Mistral · Open weight | -36.8% | | |
| 127 | Grok Code Fast 1 | xAI · Closed | -37.1% | | |
| 128 | Step 3.7 Flash | StepFun · Open weight | -37.3% | | |
| 129 | Qwen3.5 397B | Alibaba · Open weight | -37.9% | | |
| 130 | Qwen3.5 397B (Reasoning) | Alibaba · Open weight | -37.9% | | |
| 131 | MiniMax M2.5 | MiniMax · Closed | -38.9% | | |
| 132 | Mistral Large 3 | Mistral · Closed | -39.6% | | |
| 133 | GPT-4.1 | OpenAI · Closed | -39.6% | | |
| 134 | Qwen3.5-122B-A10B | Alibaba · Open weight | -41.5% | | |
| 135 | Nemotron 3 Super 100B | NVIDIA · Open weight | -41.5% | | |
| 136 | Nemotron 3 Super 120B A12B | NVIDIA · Open weight | -41.5% | | |
| 137 | DeepSeek V3 | DeepSeek · Open weight | -41.6% | | |
| 138 | Llama 4 Maverick | Meta · Open weight | -41.8% | | |
| 139 | Gemini 2.5 Flash | Google · Closed | -42.6% | | |
| 140 | DeepSeek V3.1 | DeepSeek · Open weight | -42.7% | | |
| 141 | Qwen3 Max | Alibaba · Closed | -43.5% | | |
| 142 | Qwen3.5-27B | Alibaba · Open weight | -44.0% | | |
| 143 | Trinity-Large-Thinking | Arcee AI · Open weight | -44.1% | | |
| 144 | Trinity-Large-Preview | Arcee AI · Open weight | -44.1% | | |
| 145 | Step 3.5 Flash | StepFun · Open weight | -44.2% | | |
| 146 | Nemotron Ultra 253B | NVIDIA · Open weight | -44.9% | | |
| 147 | DeepSeek V3.2 | DeepSeek · Open weight | -46.9% | | |
| 148 | Nova Pro | Amazon · Closed | -47.7% | | |
| 149 | Gemma 4 31B | Google · Open weight | -47.9% | | |
| 150 | Qwen3.5-35B-A3B | Alibaba · Open weight | -48.1% | | |
| 151 | MiMo-V2-Flash | Xiaomi · Open weight | -48.4% | | |
| 152 | MiniMax M1 80k | MiniMax · Closed | -48.5% | | |
| 153 | Claude 3 Haiku | Anthropic · Closed | -48.6% | | |
| 154 | North Mini Code | Cohere · Open weight | -48.6% | | |
| 155 | GPT-OSS 120B | OpenAI · Open weight | -49.2% | | |
| 156 | Mercury 2 | Inception · Closed | -50.7% | | |
| 157 | Gemma 4 26B A4B | Google · Open weight | -50.8% | | |
| 158 | Grok 4.1 Fast | xAI · Closed | -50.9% | | |
| 159 | Nemotron 3 Nano 30B | NVIDIA · Open weight | -51.6% | | |
| 160 | Llama 4 Scout | Meta · Open weight | -52.1% | | |
| 161 | Gemma 4 12B | Google · Open weight | -52.7% | | |
| 162 | Solar Pro 3 | Upstage · Closed | -53.4% | | |
| 163 | GPT-4.1 mini | OpenAI · Closed | -53.6% | | |
| 164 | Ultravox v0.6 Llama 3.3 70B | Fixie AI · Open weight | -54.2% | | |
| 165 | Phi-4 | Microsoft · Open weight | -55.7% | | |
| 166 | Nemotron 3 Nano Omni 30B A3B | NVIDIA · Open weight | -57.4% | | |
| 167 | GPT-4.1 nano | OpenAI · Closed | -57.6% | | |
| 168 | K-Exaone | LG AI Research · Closed | -58.0% | | |
| 169 | LFM2-24B-A2B | LiquidAI · Closed | -58.1% | | |
| 170 | Sarvam 105B | Sarvam · Open weight | -59.4% | | |
| 171 | Qwen3-Omni-30B-A3B-Thinking | Alibaba · Open weight | -61.4% | | |
| 172 | GLM-4.5-Air | Z.AI · Closed | -61.5% | | |
| 173 | Solar Pro 2 | Upstage · Closed | -62.0% | | |
| 174 | GLM-4.7-Flash | Z.AI · Open weight | -62.6% | | |
| 175 | Exaone 4.0 32B | LG AI Research · Open weight | -62.8% | | |
| 176 | GPT-OSS 20B | OpenAI · Open weight | -63.0% | | |
| 177 | Ministral 3 3B (Reasoning) | Mistral · Open weight | -64.0% | | |
| 178 | Ministral 3 3B | Mistral · Open weight | -64.0% | | |
| 179 | Ling 2.6 Flash | InclusionAI · Open weight | -66.1% | | |
| 180 | Ministral 3 14B (Reasoning) | Mistral · Open weight | -66.4% | | |
| 181 | Ministral 3 14B | Mistral · Open weight | -66.4% | | |
| 182 | Gemma 3 27B | Google · Open weight | -67.2% | | |
| 183 | Ministral 3 8B (Reasoning) | Mistral · Open weight | -68.9% | | |
| 184 | Ministral 3 8B | Mistral · Open weight | -68.9% | | |
| 185 | Qwen3-Omni-30B-A3B-Instruct | Alibaba · Open weight | -69.3% | | |
| 186 | Granite-4.0-350M | IBM · Open weight | -69.3% | | |
| 187 | Sarvam 30B | Sarvam · Open weight | -71.5% | | |
| 188 | Celeris-1 | Celeris · Closed | -71.6% | | |
| 189 | LFM2.5-1.2B-Instruct | LiquidAI · Closed | -72.1% | | |
| 190 | Granite-4.0-H-1B | IBM · Open weight | -72.3% | | |
| 191 | LFM2.5-1.2B-Thinking | LiquidAI · Closed | -79.9% | | |
| 192 | Granite-4.0-H-350M | IBM · Open weight | -80.9% | | |
| 193 | Exaone 4.0 1.2B | LG AI Research · Open weight | -82.1% | | |
| 194 | LFM2.5-VL-1.6B-Extract | LiquidAI · Open weight | -84.4% | | |