| 1 | — | Oo4-mini (high)閉源OpenAI | 69.8 | — | — | — |
| 2 | — | GGemini 2.5 Pro Preview (Mar' 25)閉源Google | 68.9 | — | — | — |
| 3 | — | Oo3閉源OpenAI | 66.8 | — | — | — |
| 4 | — | XGrok 3 mini Reasoning (high)閉源xAI | 66.7 | — | — | — |
| 5 | — | Oo3-mini (high)閉源OpenAI | 65.5 | — | — | — |
| 6 | — | Oo3-mini閉源OpenAI | 62.9 | — | — | — |
| 7 | — | AQwen3 235B A22B (Reasoning)開源Alibaba | 62.3 | — | — | — |
| 8 | — | Oo1閉源OpenAI | 61.9 | — | — | — |
| 9 | — | NLlama 3.1 Nemotron Ultra 253B v1 (Reasoning)開源NVIDIA | 60.8 | — | — | — |
| 10 | — | GGemini 2.5 Flash Preview (Reasoning)閉源Google | 60.5 | — | — | — |
| 11 | — | DDeepSeek R1開源DeepSeek | 60.2 | — | — | — |
| 12 | — | AQwen3 32B (Reasoning)開源Alibaba | 59.2 | — | — | — |
| 13 | — | AQwQ 32B開源Alibaba | 58.1 | — | — | — |
| 14 | — | AClaude 3.7 Sonnet (Extended Thinking)閉源Anthropic | 57.4 | — | — | — |
| 15 | — | AQwen3 14B (Reasoning)開源Alibaba | 55.7 | — | — | — |
| 16 | — | AQwen3 30B A3B (Reasoning)開源Alibaba | 55.6 | — | — | — |
| 17 | — | Oo1-mini閉源OpenAI | 53.8 | — | — | — |
| 18 | — | DDeepSeek V3 0324 (Mar' 25)開源DeepSeek | 53.2 | — | — | — |
| 19 | — | OGPT-4.1 mini閉源OpenAI | 52.9 | — | — | — |
| 20 | — | OGPT-4.1閉源OpenAI | 52.6 | — | — | — |
| 21 | — | GGemini 2.0 Flash Thinking Experimental (Jan '25)閉源Google | 52.3 | — | — | — |
| 22 | — | DDeepSeek R1 Distill Qwen 32B開源DeepSeek | 51.9 | — | — | — |
| 23 | 新 | AQwen3 8B (Reasoning)開源Alibaba | 51.4 | — | — | — |
| 24 | 新 | NLlama 3.3 Nemotron Super 49B v1 (Reasoning)開源NVIDIA | 51.2 | — | — | — |
| 25 | ▼ 2 | XGrok 3閉源xAI | 50.6 | — | — | — |
| 26 | ▼ 2 | MLlama 4 Maverick開源Meta | 50.5 | — | — | — |
| 27 | ▼ 2 | OGPT-4o (March 2025, chatgpt-4o-latest)閉源OpenAI | 50.3 | — | — | — |
| 28 | ▼ 2 | GGemini 2.0 Pro Experimental (Feb '25)閉源Google | 49.2 | — | — | — |
| 29 | ▼ 2 | DDeepSeek R1 Distill Qwen 14B開源DeepSeek | 49.0 | — | — | — |
| 30 | 新 | MMistral Medium 3閉源Mistral | 49.0 | — | — | — |
| 31 | ▼ 3 | GGemini 2.5 Flash Preview閉源Google | 48.8 | — | — | — |
| 32 | ▼ 4 | DDeepSeek R1 Distill Llama 70B開源DeepSeek | 48.2 | — | — | — |
| 33 | ▼ 3 | AClaude 3.7 Sonnet (Standard)閉源Anthropic | 48.2 | — | — | — |
| 34 | ▼ 3 | GGemini 2.0 Flash (Feb '25)閉源Google | 48.1 | — | — | — |
| 35 | 新 | AQwen3 4B (Reasoning)開源Alibaba | 47.3 | — | — | — |
| 36 | ▼ 4 | RReka Flash 3開源Reka AI | 47.1 | — | — | — |
| 37 | 新 | AQwen3 235B A22B開源Alibaba | 47.1 | — | — | — |
| 38 | ▼ 5 | GGemini 2.0 Flash (experimental)閉源Google | 46.4 | — | — | — |
| 39 | ▼ 5 | DDeepSeek V3 (Dec '24)開源DeepSeek | 45.6 | — | — | — |
| 40 | ▼ 5 | AQwen2.5 Max閉源Alibaba | 45.3 | — | — | — |
| 41 | ▼ 5 | GGemini 1.5 Pro (Sep '24)閉源Google | 44.6 | — | — | — |
| 42 | ▼ 5 | AClaude 3.5 Sonnet (Oct '24)閉源Anthropic | 44.4 | — | — | — |
| 43 | 新 | AQwen3 32B開源Alibaba | 43.5 | — | — | — |
| 44 | ▼ 6 | PSonar閉源Perplexity | 43.4 | — | — | — |
| 45 | ▼ 6 | MLlama 4 Scout開源Meta | 43.0 | — | — | — |
| 46 | ▼ 6 | PSonar Pro閉源Perplexity | 42.7 | — | — | — |
| 47 | ▼ 6 | AQwQ 32B-Preview開源Alibaba | 42.6 | — | — | — |
| 48 | ▼ 6 | ANova Premier閉源Amazon | 42.6 | — | — | — |
| 49 | 新 | AQwen3 30B A3B開源Alibaba | 42.6 | — | — | — |
| 50 | ▼ 7 | OGPT-4o (Nov '24)閉源OpenAI | 41.5 | — | — | — |
| 51 | ▼ 7 | GGemini 2.0 Flash-Lite (Feb '25)閉源Google | 41.4 | — | — | — |
| 52 | ▼ 7 | MLlama 3.3 Instruct 70B開源Meta | 41.1 | — | — | — |
| 53 | ▼ 7 | OGPT-4.1 nano閉源OpenAI | 41.0 | — | — | — |
| 54 | 新 | AQwen3 14B開源Alibaba | 41.0 | — | — | — |
| 55 | ▼ 8 | OGPT-4o (May '24)閉源OpenAI | 40.9 | — | — | — |
| 56 | ▼ 8 | MLlama 3.1 Instruct 405B開源Meta | 40.5 | — | — | — |
| 57 | ▼ 8 | AQwen2.5 Instruct 72B開源Alibaba | 40.4 | — | — | — |
| 58 | ▼ 8 | MMiniMax-Text-01開源MiniMax | 40.2 | — | — | — |
| 59 | ▼ 8 | MPhi-4開源Microsoft Azure | 40.2 | — | — | — |
| 60 | ▼ 8 | CCommand A開源Cohere | 40.0 | — | — | — |
| 61 | ▼ 8 | ALlama 3.1 Tulu3 405B開源Allen Institute for AI | 39.9 | — | — | — |
| 62 | ▼ 8 | NLlama 3.3 Nemotron Super 49B v1開源NVIDIA | 39.3 | — | — | — |
| 63 | ▼ 8 | XGrok 2 (Dec '24)閉源xAI | 39.2 | — | — | — |
| 64 | ▼ 8 | GGemini 1.5 Flash (Sep '24)閉源Google | 39.0 | — | — | — |
| 65 | ▼ 8 | MMistral Large 2 (Nov '24)開源Mistral | 38.3 | — | — | — |
| 66 | 新 | AQwen3 1.7B (Reasoning)開源Alibaba | 38.2 | — | — | — |
| 67 | ▼ 9 | GGemma 3 27B Instruct開源Google | 37.6 | — | — | — |
| 68 | ▼ 9 | XGrok Beta閉源xAI | 37.5 | — | — | — |
| 69 | ▼ 9 | MPixtral Large開源Mistral | 37.4 | — | — | — |
| 70 | ▼ 9 | AQwen2.5 Instruct 32B開源Alibaba | 37.4 | — | — | — |
| 71 | ▼ 9 | NLlama 3.1 Nemotron Instruct 70B開源NVIDIA | 37.3 | — | — | — |
| 72 | ▼ 9 | ANova Pro閉源Amazon | 37.1 | — | — | — |
| 73 | 新 | AQwen3 8B開源Alibaba | 36.7 | — | — | — |
| 74 | ▼ 10 | MMistral Large 2 (Jul '24)開源Mistral | 36.6 | — | — | — |
| 75 | ▼ 10 | AQwen2.5 Coder Instruct 32B開源Alibaba | 36.3 | — | — | — |
| 76 | ▼ 10 | OGPT-4o mini閉源OpenAI | 35.7 | — | — | — |
| 77 | ▼ 10 | MLlama 3.1 Instruct 70B開源Meta | 35.4 | — | — | — |
| 78 | ▼ 10 | MMistral Small 3.1開源Mistral | 35.3 | — | — | — |
| 79 | ▼ 10 | MMistral Small 3開源Mistral | 35.3 | — | — | — |
| 80 | 新 | AQwen3 4B開源Alibaba | 35.2 | — | — | — |
| 81 | ▼ 11 | AClaude 3 Opus閉源Anthropic | 35.1 | — | — | — |
| 82 | ▼ 11 | AClaude 3.5 Haiku閉源Anthropic | 34.7 | — | — | — |
| 83 | ▼ 11 | DDeepSeek R1 Distill Llama 8B開源DeepSeek | 34.0 | — | — | — |
| 84 | ▼ 11 | GGemma 3 12B Instruct開源Google | 33.8 | — | — | — |
| 85 | ▼ 11 | GGemini 1.5 Pro (May '24)閉源Google | 33.7 | — | — | — |
| 86 | ▼ 11 | AQwen Turbo閉源Alibaba | 33.6 | — | — | — |
| 87 | ▼ 11 | MLlama 3.2 Instruct 90B (Vision)開源Meta | 33.4 | — | — | — |
| 88 | ▼ 11 | AQwen2 Instruct 72B開源Alibaba | 32.6 | — | — | — |
| 89 | ▼ 11 | ANova Lite閉源Amazon | 32.5 | — | — | — |
| 90 | ▼ 11 | GGemini 1.5 Flash-8B閉源Google | 30.8 | — | — | — |
| 91 | ▼ 11 | NDeepHermes 3 - Mistral 24B Preview開源Nous Research | 30.0 | — | — | — |
| 92 | ▼ 11 | AJamba 1.5 Large開源AI21 Labs | 29.3 | — | — | — |
| 93 | ▼ 11 | NHermes 3 - Llama-3.1 70B開源Nous Research | 29.1 | — | — | — |
| 94 | ▼ 11 | AJamba 1.6 Large開源AI21 Labs | 28.8 | — | — | — |
| 95 | ▼ 11 | GGemini 1.5 Flash (May '24)閉源Google | 28.5 | — | — | — |
| 96 | ▼ 11 | ANova Micro閉源Amazon | 28.3 | — | — | — |
| 97 | ▼ 11 | 0Yi-Large閉源01.AI | 27.9 | — | — | — |
| 98 | ▼ 11 | AClaude 3 Sonnet閉源Anthropic | 27.8 | — | — | — |
| 99 | ▼ 11 | MCodestral (Jan '25)閉源Mistral | 27.6 | — | — | — |
| 100 | ▼ 11 | MLlama 3 Instruct 70B開源Meta | 27.5 | — | — | — |
| 101 | ▼ 11 | MMistral Small (Sep '24)開源Mistral | 27.4 | — | — | — |
| 102 | ▼ 11 | MPhi-4 Multimodal Instruct開源Microsoft Azure | 26.9 | — | — | — |
| 103 | ▼ 11 | AQwen2.5 Coder Instruct 7B開源Alibaba | 26.7 | — | — | — |
| 104 | ▼ 11 | MMistral Large (Feb '24)閉源Mistral | 26.4 | — | — | — |
| 105 | ▼ 11 | MMixtral 8x22B Instruct開源Mistral | 26.2 | — | — | — |
| 106 | ▼ 11 | MPhi-4 Mini Instruct開源Microsoft Azure | 26.0 | — | — | — |
| 107 | 新 | AQwen3 1.7B開源Alibaba | 25.0 | — | — | — |
| 108 | ▼ 12 | MPhi-3 Medium Instruct 14B開源Microsoft Azure | 24.5 | — | — | — |
| 109 | ▼ 12 | GGemma 3 4B Instruct開源Google | 24.5 | — | — | — |
| 110 | ▼ 12 | AClaude 2.1閉源Anthropic | 24.1 | — | — | — |
| 111 | ▼ 12 | MLlama 3.1 Instruct 8B開源Meta | 23.7 | — | — | — |
| 112 | ▼ 12 | MPixtral 12B (2409)開源Mistral | 23.4 | — | — | — |
| 113 | 新 | AQwen3 0.6B (Reasoning)開源Alibaba | 23.2 | — | — | — |
| 114 | ▼ 13 | MMistral Small (Feb '24)閉源Mistral | 22.9 | — | — | — |
| 115 | ▼ 13 | MMistral Medium閉源Mistral | 22.8 | — | — | — |
| 116 | ▼ 13 | MMinistral 8B開源Mistral | 22.3 | — | — | — |
| 117 | ▼ 13 | GGemma 2 9B開源Google | 22.2 | — | — | — |
| 118 | ▼ 13 | MPhi-3 Mini Instruct 3.8B開源Microsoft Azure | 22.1 | — | — | — |
| 119 | ▼ 13 | LLFM 40B閉源Liquid AI | 21.7 | — | — | — |
| 120 | ▼ 13 | CCommand-R+ (Aug '24)開源Cohere | 21.5 | — | — | — |
| 121 | ▼ 13 | MLlama 3 Instruct 8B開源Meta | 21.5 | — | — | — |
| 122 | ▼ 13 | GGemini 1.0 Pro閉源Google | 20.6 | — | — | — |
| 123 | ▼ 13 | MCodestral (May '24)開源Mistral | 20.4 | — | — | — |
| 124 | ▼ 13 | CAya Expanse 32B開源Cohere | 20.1 | — | — | — |
| 125 | ▼ 13 | MLlama 2 Chat 13B開源Meta | 19.9 | — | — | — |
| 126 | ▼ 13 | CCommand-R+ (Apr '24)開源Cohere | 19.9 | — | — | — |
| 127 | ▼ 13 | DDBRX Instruct開源Databricks | 19.8 | — | — | — |
| 128 | ▼ 13 | MMinistral 3B閉源Mistral | 19.6 | — | — | — |
| 129 | ▼ 13 | MMistral NeMo開源Mistral | 19.6 | — | — | — |
| 130 | ▼ 13 | MLlama 3.2 Instruct 3B開源Meta | 19.5 | — | — | — |
| 131 | ▼ 13 | DDeepSeek R1 Distill Qwen 1.5B開源DeepSeek | 19.2 | — | — | — |
| 132 | ▼ 13 | AJamba 1.5 Mini開源AI21 Labs | 18.4 | — | — | — |
| 133 | ▼ 13 | AJamba 1.6 Mini開源AI21 Labs | 17.7 | — | — | — |
| 134 | ▼ 13 | MMixtral 8x7B Instruct開源Mistral | 17.0 | — | — | — |
| 135 | 新 | AQwen3 0.6B開源Alibaba | 16.7 | — | — | — |
| 136 | ▼ 14 | NDeepHermes 3 - Llama-3.1 8B Preview開源Nous Research | 16.2 | — | — | — |
| 137 | ▼ 14 | CAya Expanse 8B開源Cohere | 16.0 | — | — | — |
| 138 | ▼ 14 | CCommand-R (Aug '24)開源Cohere | 14.8 | — | — | — |
| 139 | ▼ 14 | CCommand-R (Mar '24)開源Cohere | 14.7 | — | — | — |
| 140 | ▼ 14 | MCodestral-Mamba開源Mistral | 13.9 | — | — | — |
| 141 | ▼ 14 | MMistral 7B Instruct開源Mistral | 10.1 | — | — | — |
| 142 | ▼ 14 | MLlama 3.2 Instruct 1B開源Meta | 9.7 | — | — | — |
| 143 | ▼ 14 | MLlama 2 Chat 7B開源Meta | 8.3 | — | — | — |