| 1 | GPT-5.5 | OpenAI | Closed | 74.1% | |
| 2 | Claude Opus 4.7 (Adaptive) | Anthropic | Closed | 71.3% | |
| 3 | Gemini 3.5 Flash | Google | Closed | 70.3% | |
| 4 | GPT-5.4 | OpenAI | Closed | 68.0% | |
| 5 | Claude Opus 4.6 (Adaptive) | Anthropic | Closed | 67.6% | |
| 6 | MiMo-V2.5-Pro | Xiaomi | Closed | 67.4% | |
| 7 | DeepSeek V4 Pro (Max) | DeepSeek | Open | 67.2% | |
| 8 | GLM-5.1 | Z.AI | Open | 67.0% | |
| 9 | DeepSeek V4 Pro (High) | DeepSeek | Open | 66.7% | |
| 10 | Qwen3.7 Max | Alibaba | Closed | 66.6% | |
| 11 | GLM-5-Turbo | Z.AI | Closed | 66.1% | |
| 12 | Kimi K2.6 | Moonshot AI | Open | 66.0% | |
| 13 | Grok 4.3 | xAI | Closed | 65.9% | |
| 14 | Qwen 3.6 Max (preview) | Alibaba | Closed | 64.8% | |
| 15 | Claude Opus 4.7 | Anthropic | Closed | 64.6% | |
| 16 | Claude Opus 4.6 | Anthropic | Closed | 64.2% | |
| 17 | GLM-5 | Z.AI | Open | 63.1% | |
| 18 | Qwen3.6-27B | Alibaba | Open | 62.9% | |
| 19 | MiMo-V2-Pro | Xiaomi | Closed | 62.8% | |
| 20 | DeepSeek V4 Flash (High) | DeepSeek | Open | 62.3% | |
| 21 | Muse Spark | Meta | Closed | 62.0% | |
| 22 | Qwen3.6 Plus | Alibaba | Closed | 61.7% | |
| 23 | Claude Sonnet 4.6 | Anthropic | Closed | 61.6% | |
| 24 | MiniMax M2.7 | MiniMax | Open | 61.5% | |
| 25 | DeepSeek V4 Flash (Max) | DeepSeek | Open | 61.3% | |
| 26 | GLM-5V-Turbo | Z.AI | Closed | 61.1% | |
| 27 | GPT-5.3 Codex | OpenAI | Closed | 60.5% | |
| 28 | GPT-5.2 | OpenAI | Closed | 60.2% | |
| 29 | Claude Opus 4.5 Thinking | Anthropic | Closed | 59.6% | |
| 30 | Claude Opus 4.5 | Anthropic | Closed | 59.2% | |
| 31 | Gemini 3.1 Pro | Google | Closed | 59.1% | |
| 32 | Kimi K2.5 (Reasoning) | Moonshot AI | Closed | 58.9% | |
| 33 | Kimi K2.5 | Moonshot AI | Open | 58.9% | |
| 34 | GPT-5.4 mini | OpenAI | Closed | 58.9% | |
| 35 | MiMo-V2-Omni | Xiaomi | Closed | 58.6% | |
| 36 | Qwen3.6-35B-A3B | Alibaba | Open | 58.3% | |
| 37 | GPT-5.2-Codex | OpenAI | Closed | 56.5% | |
| 38 | Qwen3.5 397B (Reasoning) | Alibaba | Open | 55.8% | |
| 39 | Hy3 Preview | Tencent | Open | 55.7% | |
| 40 | GLM-4.7 | Z.AI | Open | 55.0% | |
| 41 | GPT-5 (high) | OpenAI | Closed | 54.6% | |
| 42 | Qwen3.5-27B | Alibaba | Open | 54.6% | |
| 43 | Qwen3.5 397B | Alibaba | Open | 53.3% | |
| 44 | Mistral Medium 3.5 128B | Mistral | Open | 53.2% | |
| 45 | Qwen3.5-122B-A10B | Alibaba | Open | 53.0% | |
| 46 | Gemini 3 Pro | Google | Closed | 52.0% | |
| 47 | GPT-5.1 | OpenAI | Closed | 51.3% | |
| 48 | GPT-5.1-Codex-Max | OpenAI | Closed | 50.7% | |
| 49 | GPT-5.1-Codex | OpenAI | Closed | 50.7% | |
| 50 | Grok 4.1 Fast (Reasoning) | xAI | Closed | 49.3% | |
| 51 | GPT-5.4 nano | OpenAI | Closed | 47.6% | |
| 52 | MiMo-V2-Flash | Xiaomi | Open | 47.3% | |
| 53 | GPT-5 (medium) | OpenAI | Closed | 45.8% | |
| 54 | Qwen3.5-35B-A3B | Alibaba | Open | 44.1% | |
| 55 | Qwen3 Max | Alibaba | Closed | 43.0% | |
| 56 | GLM-4.6 | Z.AI | Open | 42.9% | |
| 57 | Trinity-Large-Preview | Arcee AI | Open | 42.6% | |
| 58 | Trinity-Large-Thinking | Arcee AI | Open | 42.6% | |
| 59 | Grok 4 | xAI | Closed | 41.5% | |
| 60 | Gemma 4 31B | Google | Open | 40.9% | |
| 61 | Command A+ | Cohere | Open | 40.9% | |
| 62 | DeepSeek V3.2 | DeepSeek | Open | 39.8% | |
| 63 | Grok 4 Fast (Reasoning) | xAI | Closed | 39.5% | |
| 64 | Claude 4 Sonnet | Anthropic | Closed | 39.2% | |
| 65 | K-Exaone | LG AI Research | Closed | 38.1% | |
| 66 | Ling 2.6 Flash | InclusionAI | Open | 38.1% | |
| 67 | GPT-OSS 120B | OpenAI | Open | 37.9% | |
| 68 | o3 | OpenAI | Closed | 36.1% | |
| 69 | Grok Code Fast 1 | xAI | Closed | 35.6% | |
| 70 | Gemini 3 Flash | Google | Closed | 35.0% | |
| 71 | Grok 4.1 Fast | xAI | Closed | 33.0% | |
| 72 | Gemini 2.5 Pro | Google | Closed | 32.7% | |
| 73 | Gemma 4 26B A4B | Google | Open | 32.1% | |
| 74 | DeepSeek V3.1 | DeepSeek | Open | 31.9% | |
| 75 | o1 | OpenAI | Closed | 31.1% | |
| 76 | GPT-OSS 20B | OpenAI | Open | 27.6% | |
| 77 | GPT-4.1 | OpenAI | Closed | 27.3% | |
| 78 | Mistral Small 4 (Reasoning) | Mistral | Open | 25.9% | |
| 79 | Mistral Small 4 | Mistral | Open | 25.9% | |
| 80 | Gemini 3.1 Flash-Lite | Google | Closed | 25.7% | |
| 81 | GPT-4.1 mini | OpenAI | Closed | 25.1% | |
| 82 | Sarvam 105B | Sarvam | Open | 24.7% | |
| 83 | Kimi K2 | Moonshot AI | Closed | 24.3% | |
| 84 | Nemotron 3 Nano Omni 30B A3B | NVIDIA | Open | 23.9% | |
| 85 | Mistral Large 3 | Mistral | Closed | 21.7% | |
| 86 | GLM-4.5-Air | Z.AI | Closed | 21.0% | |
| 87 | DeepSeek-R1 | DeepSeek | Open | 20.8% | |
| 88 | DeepSeek V3.1 (Reasoning) | DeepSeek | Open | 18.9% | |
| 89 | Gemini 2.5 Flash | Google | Closed | 15.0% | |
| 90 | Mistral Medium 3 | Mistral | Closed | 13.7% | |
| 91 | Solar Pro 2 | Upstage | Closed | 12.7% | |
| 92 | Sarvam 30B | Sarvam | Open | 11.5% | |
| 93 | Mistral Large 2 | Mistral | Closed | 10.2% | |
| 94 | DeepSeek V3 | DeepSeek | Open | 8.8% | |
| 95 | Nemotron 3 Nano 30B | NVIDIA | Open | 8.5% | |
| 96 | GPT-4o | OpenAI | Closed | 8.4% | |
| 97 | Granite-4.0-1B | IBM | Open | 7.6% | |
| 98 | Llama 4 Maverick | Meta | Open | 7.2% | |
| 99 | Claude 3 Haiku | Anthropic | Closed | 7.0% | |
| 100 | Gemma 4 E4B | Google | Open | 6.9% | |
| 101 | Gemma 4 E2B | Google | Open | 6.9% | |
| 102 | Exaone 4.0 1.2B | LG AI Research | Open | 6.8% | |
| 103 | Granite-4.0-H-1B | IBM | Open | 6.5% | |
| 104 | Llama 3.1 405B | Meta | Open | 6.3% | |
| 105 | GPT-4.1 nano | OpenAI | Closed | 5.8% | |
| 106 | Llama 4 Scout | Meta | Open | 5.2% | |
| 107 | Granite-4.0-H-350M | IBM | Open | 4.9% | |
| 108 | Nova Pro | Amazon | Closed | 4.7% | |
| 109 | Granite-4.0-350M | IBM | Open | 4.4% | |
| 110 | Nemotron Ultra 253B | NVIDIA | Open | 3.8% | |
| 111 | Gemma 3 27B | Google | Open | 3.5% | |
| 112 | Exaone 4.0 32B | LG AI Research | Open | 1.4% | |
| 113 | Phi-4 | Microsoft | Open | 0.0% | |