AI 大模型排名 ArtificialAnalysis AI 大模型排行榜
🔍 信息查询
32
次浏览
100%
有帮助
·
1 人反馈
AI 大模型排名 Artificial Analysis AI 大模型排行榜，综合对超过 100 个 AI 模型（LLM）的性能进行了比较和排名，评估指标包括智能程度、价格以及常见AI基准测试的结果。
AI 大模型排行榜数据中心
| 排名 | 模型名称 | 综合指数 ▼ | 编程 | 价格 ($/1M) |
|---|---|---|---|---|
| 1 | GPT-5.4 (xhigh) | 57.2 | 57.3 | $5.625 |
| 2 | Gemini 3.1 Pro Preview | 57.2 | 55.5 | $4.5 |
| 3 | GPT-5.3 Codex (xhigh) | 54 | 53.1 | $4.813 |
| 4 | Claude Opus 4.6 (Adaptive Reasoning, Max Effort) | 53 | 48.1 | $10 |
| 5 | Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort) | 51.7 | 50.9 | $6 |
| 6 | GPT-5.2 (xhigh) | 51.3 | 48.7 | $4.813 |
| 7 | GLM-5 (Reasoning) | 49.8 | 44.2 | $1.55 |
| 8 | Claude Opus 4.5 (Reasoning) | 49.7 | 47.8 | $10 |
| 9 | MiniMax-M2.7 | 49.6 | 41.9 | $0.525 |
| 10 | MiMo-V2-Pro | 49.2 | 41.4 | $1.5 |
| 11 | GPT-5.2 Codex (xhigh) | 49 | 43 | $4.813 |
| 12 | Grok 4.20 Beta 0309 (Reasoning) | 48.5 | 42.2 | $3 |
| 13 | Gemini 3 Pro Preview (high) | 48.4 | 46.5 | $4.5 |
| 14 | GPT-5.4 mini (xhigh) | 48.1 | 51.5 | $1.688 |
| 15 | GPT-5.1 (high) | 47.7 | 44.7 | $3.438 |
| 16 | Kimi K2.5 (Reasoning) | 46.8 | 39.5 | $1.2 |
| 17 | GLM-5-Turbo | 46.8 | 36.8 | $0 |
| 18 | GPT-5.2 (medium) | 46.6 | 44.2 | $4.813 |
| 19 | Claude Opus 4.6 (Non-reasoning, High Effort) | 46.5 | 47.6 | $10 |
| 20 | Gemini 3 Flash Preview (Reasoning) | 46.4 | 42.6 | $1.125 |
| 21 | Qwen3.5 397B A17B (Reasoning) | 45 | 41.3 | $1.35 |
| 22 | GPT-5 (high) | 44.6 | 36 | $3.438 |
| 23 | GPT-5 Codex (high) | 44.6 | 38.9 | $3.438 |
| 24 | GPT-5.4 nano (xhigh) | 44.4 | 43.9 | $0.463 |
| 25 | Claude Sonnet 4.6 (Non-reasoning, High Effort) | 44.4 | 46.4 | $6 |
| 26 | MiMo-V2-Omni | 43.4 | 35.5 | $0 |
| 27 | GPT-5.1 Codex (high) | 43.1 | 36.6 | $3.438 |
| 28 | Claude Opus 4.5 (Non-reasoning) | 43.1 | 42.9 | $10 |
| 29 | Claude 4.5 Sonnet (Reasoning) | 43 | 38.6 | $6 |
| 30 | Claude Sonnet 4.6 (Non-reasoning, Low Effort) | 42.6 | 43 | $6 |
| 31 | Qwen3.5 27B (Reasoning) | 42.1 | 34.9 | $0.825 |
| 32 | GLM-4.7 (Reasoning) | 42.1 | 36.3 | $1 |
| 33 | GPT-5 (medium) | 42 | 39 | $3.438 |
| 34 | Claude 4.1 Opus (Reasoning) | 42 | 36.5 | $30 |
| 35 | MiniMax-M2.5 | 41.9 | 37.4 | $0.525 |
| 36 | DeepSeek V3.2 (Reasoning) | 41.7 | 36.7 | $0.315 |
| 37 | Qwen3.5 122B A10B (Reasoning) | 41.6 | 34.7 | $1.1 |
| 38 | MiMo-V2-Flash (Feb 2026) | 41.5 | 33.5 | $0.15 |
| 39 | Grok 4 | 41.5 | 40.5 | $6 |
| 40 | Gemini 3 Pro Preview (low) | 41.3 | 39.4 | $4.5 |
| 41 | GPT-5 mini (high) | 41.2 | 35.3 | $0.688 |
| 42 | Kimi K2 Thinking | 40.9 | 34.8 | $1.075 |
| 43 | o3-pro | 40.7 | - | $35 |
| 44 | GLM-5 (Non-reasoning) | 40.6 | 39 | $1.55 |
| 45 | Qwen3.5 397B A17B (Non-reasoning) | 40.1 | 37.4 | $1.35 |
| 46 | Qwen3 Max Thinking | 39.9 | 30.5 | $2.4 |
| 47 | MiniMax-M2.1 | 39.4 | 32.8 | $0.525 |
| 48 | GPT-5 (low) | 39.2 | 30.7 | $3.438 |
| 49 | MiMo-V2-Flash (Reasoning) | 39.2 | 31.8 | $0.15 |
| 50 | Claude 4 Opus (Reasoning) | 39 | 34 | $30 |
| 51 | GPT-5 mini (medium) | 38.9 | 32.9 | $0.688 |
| 52 | Claude 4 Sonnet (Reasoning) | 38.7 | 34.1 | $6 |
| 53 | GPT-5.1 Codex mini (high) | 38.6 | 36.4 | $0.688 |
| 54 | Grok 4.1 Fast (Reasoning) | 38.6 | 30.9 | $0.275 |
| 55 | o3 | 38.4 | 38.4 | $3.5 |
| 56 | GPT-5.4 nano (medium) | 38.1 | 35 | $0.463 |
| 57 | Step 3.5 Flash | 37.8 | 31.6 | $0.15 |
| 58 | GPT-5.4 mini (medium) | 37.7 | 37.5 | $1.688 |
| 59 | Kimi K2.5 (Non-reasoning) | 37.3 | 25.8 | $1.2 |
| 60 | Qwen3.5 27B (Non-reasoning) | 37.2 | 33.4 | $0.825 |
| 61 | Claude 4.5 Haiku (Reasoning) | 37.1 | 32.6 | $2 |
| 62 | Qwen3.5 35B A3B (Reasoning) | 37.1 | 30.3 | $0.688 |
| 63 | Claude 4.5 Sonnet (Non-reasoning) | 37.1 | 33.5 | $6 |
| 64 | MiniMax-M2 | 36.1 | 29.2 | $0.525 |
| 65 | NVIDIA Nemotron 3 Super 120B A12B (Reasoning) | 36 | 31.2 | $0.412 |
| 66 | KAT-Coder-Pro V1 | 36 | 18.3 | $0.525 |
| 67 | Claude 4.1 Opus (Non-reasoning) | 36 | - | $30 |
| 68 | Qwen3.5 122B A10B (Non-reasoning) | 35.9 | 31.6 | $1.1 |
| 69 | Nova 2.0 Pro Preview (medium) | 35.7 | 30.4 | $3.438 |
| 70 | GPT-5.4 (Non-reasoning) | 35.4 | 41 | $5.625 |
| 71 | Grok 4 Fast (Reasoning) | 35.1 | 27.4 | $0.275 |
| 72 | Gemini 3 Flash Preview (Non-reasoning) | 35 | 37.8 | $1.125 |
| 73 | Claude 3.7 Sonnet (Reasoning) | 34.7 | 27.6 | $6 |
| 74 | Gemini 2.5 Pro | 34.6 | 31.9 | $3.438 |
| 75 | GLM-4.7 (Non-reasoning) | 34.2 | 32 | $0.938 |
| 76 | DeepSeek V3.1 Terminus (Reasoning) | 33.9 | 33.7 | $0.8 |
| 77 | GPT-5.2 (Non-reasoning) | 33.6 | 34.7 | $4.813 |
| 78 | Gemini 3.1 Flash-Lite Preview | 33.5 | 30.1 | $0.563 |
| 79 | Doubao Seed Code | 33.5 | 31.3 | $0 |
| 80 | gpt-oss-120B (high) | 33.3 | 28.6 | $0.263 |
| 81 | o4-mini (high) | 33.1 | 25.6 | $1.925 |
| 82 | Claude 4 Opus (Non-reasoning) | 33 | - | $30 |
| 83 | Claude 4 Sonnet (Non-reasoning) | 33 | 30.6 | $6 |
| 84 | DeepSeek V3.2 Exp (Reasoning) | 32.9 | 33.3 | $0.315 |
| 85 | Mercury 2 | 32.8 | 30.6 | $0.375 |
| 86 | GLM-4.6 (Reasoning) | 32.5 | 29.5 | $0.981 |
| 87 | Qwen3 Max Thinking (Preview) | 32.5 | 24.5 | $2.4 |
| 88 | Qwen3.5 9B (Reasoning) | 32.4 | 25.3 | $0.113 |
| 89 | DeepSeek V3.2 (Non-reasoning) | 32.1 | 34.6 | $0.315 |
| 90 | Grok 3 mini Reasoning (high) | 32.1 | 25.2 | $0.35 |
| 91 | K-EXAONE (Reasoning) | 32.1 | 27 | $0 |
| 92 | Nova 2.0 Pro Preview (low) | 31.9 | 24.5 | $3.438 |
| 93 | Qwen3 Max | 31.4 | 26.4 | $2.4 |
| 94 | Claude 4.5 Haiku (Non-reasoning) | 31.1 | 29.6 | $2 |
| 95 | Gemini 2.5 Flash Preview (Sep '25) (Reasoning) | 31.1 | 24.6 | $0 |
| 96 | Kimi K2 0905 | 30.9 | 25.9 | $1.137 |
| 97 | o1 | 30.8 | 20.5 | $26.25 |
| 98 | Claude 3.7 Sonnet (Non-reasoning) | 30.8 | 26.7 | $6 |
| 99 | Qwen3.5 35B A3B (Non-reasoning) | 30.7 | 16.8 | $0.688 |
| 100 | MiMo-V2-Flash (Non-reasoning) | 30.4 | 25.8 | $0.15 |
| 101 | Gemini 2.5 Pro Preview (Mar' 25) | 30.3 | 46.7 | $0 |
| 102 | GLM-4.6 (Non-reasoning) | 30.2 | 30.2 | $1 |
| 103 | GLM-4.7-Flash (Reasoning) | 30.1 | 25.9 | $0.152 |
| 104 | Grok 4.20 Beta 0309 (Non-reasoning) | 29.7 | 25.4 | $3 |
| 105 | Nova 2.0 Lite (medium) | 29.7 | 23.9 | $0.85 |
| 106 | Gemini 2.5 Pro Preview (May' 25) | 29.5 | - | $3.438 |
| 107 | Qwen3 235B A22B 2507 (Reasoning) | 29.5 | 23.2 | $2.625 |
| 108 | DeepSeek V3.2 Speciale | 29.4 | 37.9 | $0 |
| 109 | ERNIE 5.0 Thinking Preview | 29.1 | 29.2 | $0 |
| 110 | Grok Code Fast 1 | 28.7 | 23.7 | $0.525 |
| 111 | DeepSeek V3.1 Terminus (Non-reasoning) | 28.5 | 31.9 | $0.626 |
| 112 | DeepSeek V3.2 Exp (Non-reasoning) | 28.4 | 30 | $0.315 |
| 113 | Qwen3 Coder Next | 28.3 | 22.9 | $0.6 |
| 114 | Apriel-v1.5-15B-Thinker | 28.3 | 18.7 | $0 |
| 115 | DeepSeek V3.1 (Non-reasoning) | 28.1 | 28.4 | $0.84 |
| 116 | Nova 2.0 Omni (medium) | 28 | 15.1 | $0.85 |
| 117 | DeepSeek V3.1 (Reasoning) | 27.7 | 29.7 | $0.875 |
| 118 | Apriel-v1.6-15B-Thinker | 27.6 | 22 | $0 |
| 119 | Qwen3 VL 235B A22B (Reasoning) | 27.6 | 20.9 | $2.625 |
| 120 | GPT-5.1 (Non-reasoning) | 27.4 | 27.3 | $3.438 |
| 121 | Qwen3.5 9B (Non-reasoning) | 27.3 | 21.4 | $0 |
| 122 | Magistral Medium 1.2 | 27.1 | 21.7 | $2.75 |
| 123 | DeepSeek R1 0528 (May '25) | 27.1 | 24 | $2.362 |
| 124 | Qwen3.5 4B (Reasoning) | 27.1 | 17.5 | $0 |
| 125 | Gemini 2.5 Flash (Reasoning) | 27 | 22.2 | $0.85 |
| 126 | Mistral Small 4 (Reasoning) | 26.9 | 24.3 | $0.263 |
| 127 | GPT-5 nano (high) | 26.8 | 20.3 | $0.138 |
| 128 | Qwen3 Next 80B A3B (Reasoning) | 26.7 | 19.5 | $1.875 |
| 129 | GLM-4.5 (Reasoning) | 26.4 | 26.3 | $0.843 |
| 130 | GPT-4.1 | 26.3 | 21.8 | $3.5 |
| 131 | Kimi K2 | 26.3 | 22.1 | $1.002 |
| 132 | Qwen3 Max (Preview) | 26.1 | 25.5 | $2.4 |
| 133 | GPT-5 nano (medium) | 25.9 | 22.9 | $0.138 |
| 134 | o3-mini | 25.9 | 17.9 | $1.925 |
| 135 | o1-pro | 25.8 | - | $262.5 |
| 136 | Gemini 2.5 Flash Preview (Sep '25) (Non-reasoning) | 25.7 | 22.1 | $0 |
| 137 | o3-mini (high) | 25.2 | 17.3 | $1.925 |
| 138 | Grok 3 | 25.2 | 19.8 | $6 |
| 139 | Seed-OSS-36B-Instruct | 25.2 | 16.7 | $0.3 |
| 140 | Qwen3 235B A22B 2507 Instruct | 25 | 22.1 | $1.225 |
| 141 | Qwen3 Coder 480B A35B Instruct | 24.8 | 24.6 | $3 |
| 142 | Qwen3 VL 32B (Reasoning) | 24.7 | 14.5 | $2.625 |
| 143 | Nova 2.0 Lite (low) | 24.6 | 13.6 | $0.85 |
| 144 | Sonar Reasoning Pro | 24.6 | - | $0 |
| 145 | gpt-oss-120B (low) | 24.5 | 15.5 | $0.263 |
| 146 | gpt-oss-20B (high) | 24.5 | 18.5 | $0.094 |
| 147 | GPT-5.4 nano (Non-Reasoning) | 24.4 | 27.9 | $0.463 |
| 148 | MiniMax M1 80k | 24.4 | 14.5 | $0.963 |
| 149 | NVIDIA Nemotron 3 Nano 30B A3B (Reasoning) | 24.3 | 19 | $0.105 |
| 150 | Gemini 2.5 Flash Preview (Reasoning) | 24.3 | - | $0 |
| 151 | K2 Think V2 | 24.1 | 15.5 | $0 |
| 152 | LongCat Flash Lite | 23.9 | 16.5 | $0 |
| 153 | GPT-5 (minimal) | 23.9 | 25.1 | $3.438 |
| 154 | HyperCLOVA X SEED Think (32B) | 23.7 | 17.5 | $0 |
| 155 | o1-preview | 23.7 | 34 | $28.875 |
| 156 | Grok 4.1 Fast (Non-reasoning) | 23.6 | 19.5 | $0.275 |
| 157 | K-EXAONE (Non-reasoning) | 23.4 | 13.5 | $0 |
| 158 | GLM-4.6V (Reasoning) | 23.4 | 19.7 | $0.45 |
| 159 | GPT-5.4 mini (Non-Reasoning) | 23.3 | 25.3 | $1.688 |
| 160 | Nova 2.0 Omni (low) | 23.2 | 13.9 | $0.85 |
| 161 | GLM-4.5-Air | 23.2 | 23.8 | $0.425 |
| 162 | Nova 2.0 Pro Preview (Non-reasoning) | 23.1 | 20.5 | $3.438 |
| 163 | Mi:dm K 2.5 Pro | 23.1 | 12.6 | $0 |
| 164 | Grok 4 Fast (Non-reasoning) | 23.1 | 19 | $0.275 |
| 165 | GPT-4.1 mini | 22.9 | 18.5 | $0.7 |
| 166 | Mistral Large 3 | 22.8 | 22.7 | $0.75 |
| 167 | Ring-1T | 22.8 | 16.8 | $0 |
| 168 | Qwen3.5 4B (Non-reasoning) | 22.6 | 13.7 | $0 |
| 169 | Qwen3 30B A3B 2507 (Reasoning) | 22.4 | 14.7 | $0.75 |
| 170 | DeepSeek V3 0324 | 22.3 | 22 | $1.25 |
| 171 | INTELLECT-3 | 22.2 | 19.1 | $0 |
| 172 | GLM-4.7-Flash (Non-reasoning) | 22.1 | 11 | $0.152 |
| 173 | Devstral 2 | 22 | 23.7 | $0 |
| 174 | GPT-5 (ChatGPT) | 21.8 | 21.2 | $3.438 |
| 175 | Solar Open 100B (Reasoning) | 21.7 | 10.5 | $0 |
| 176 | Gemini 2.5 Flash-Lite Preview (Sep '25) (Reasoning) | 21.6 | 18.1 | $0.175 |
| 177 | Grok 3 Reasoning Beta | 21.6 | - | $0 |
| 178 | Mistral Medium 3.1 | 21.3 | 18.3 | $0.8 |
| 179 | MiniMax M1 40k | 20.9 | 14.1 | $0 |
| 180 | gpt-oss-20B (low) | 20.8 | 14.4 | $0.094 |
| 181 | Qwen3 VL 235B A22B Instruct | 20.8 | 16.5 | $1.225 |
| 182 | GPT-5 mini (minimal) | 20.7 | 21.9 | $0.688 |
| 183 | K2-V2 (high) | 20.6 | 16.1 | $0 |
| 184 | Gemini 2.5 Flash (Non-reasoning) | 20.6 | 17.8 | $0.85 |
| 185 | o1-mini | 20.4 | - | $0 |
| 186 | Qwen3 Next 80B A3B Instruct | 20.1 | 15.3 | $0.875 |
| 187 | Tri-21B-think Preview | 20 | 7.4 | $0 |
| 188 | GPT-4.5 (Preview) | 20 | - | $0 |
| 189 | Qwen3 Coder 30B A3B Instruct | 20 | 19.4 | $0.9 |
| 190 | Qwen3 235B A22B (Reasoning) | 19.8 | 17.4 | $2.625 |
| 191 | QwQ 32B | 19.7 | - | $0.745 |
| 192 | Qwen3 VL 30B A3B (Reasoning) | 19.7 | 13.1 | $0.75 |
| 193 | Gemini 2.0 Flash Thinking Experimental (Jan '25) | 19.6 | 24.1 | $0 |
| 194 | Devstral Small 2 | 19.5 | 20.7 | $0 |
| 195 | Gemini 2.5 Flash-Lite Preview (Sep '25) (Non-reasoning) | 19.4 | 14.5 | $0.175 |
| 196 | Motif-2-12.7B-Reasoning | 19.1 | 11.9 | $0 |
| 197 | Nova Premier | 19 | 13.8 | $5 |
| 198 | Ling-1T | 19 | 18.8 | $0 |
| 199 | Mistral Medium 3 | 18.8 | 13.6 | $0.8 |
| 200 | Magistral Medium 1 | 18.8 | 16 | $0 |
| 201 | DeepSeek R1 (Jan '25) | 18.8 | 15.9 | $2.362 |
| 202 | Solar Pro 2 (Preview) (Reasoning) | 18.8 | - | $0 |
| 203 | Llama Nemotron Super 49B v1.5 (Reasoning) | 18.7 | 15.2 | $0.175 |
| 204 | K2-V2 (medium) | 18.7 | 14 | $0 |
| 205 | Claude 3.5 Haiku | 18.7 | 10.7 | $1.6 |
| 206 | Devstral Medium | 18.7 | 15.9 | $0.8 |
| 207 | Mistral Small 4 (Non-reasoning) | 18.6 | 16.4 | $0.263 |
| 208 | Hermes 4 - Llama-3.1 405B (Reasoning) | 18.6 | 16 | $1.5 |
| 209 | Tri-21B-Think | 18.6 | 6.3 | $0 |
| 210 | GPT-4o (Aug '24) | 18.6 | 16.6 | $4.375 |
| 211 | GPT-4o (March 2025, chatgpt-4o-latest) | 18.6 | - | $0 |
| 212 | Llama 3.3 Nemotron Super 49B v1 (Reasoning) | 18.5 | 9.4 | $0 |
| 213 | Gemini 2.0 Flash (Feb '25) | 18.5 | 13.6 | $0.263 |
| 214 | Llama 4 Maverick | 18.4 | 15.6 | $0.487 |
| 215 | Magistral Small 1.2 | 18.2 | 14.8 | $0.75 |
| 216 | Sarvam 105B (high) | 18.2 | 9.8 | $0 |
| 217 | Qwen3 4B 2507 (Reasoning) | 18.2 | 9.5 | $0 |
| 218 | Gemini 2.0 Pro Experimental (Feb '25) | 18.1 | 25.5 | $0 |
| 219 | Nova 2.0 Lite (Non-reasoning) | 18 | 12.5 | $0.85 |
| 220 | Claude 3 Opus | 18 | 19.5 | $30 |
| 221 | Devstral Small (May '25) | 18 | 12.2 | $0.075 |
| 222 | Sonar Reasoning | 17.9 | - | $0 |
| 223 | Gemini 2.5 Flash Preview (Non-reasoning) | 17.8 | - | $0 |
| 224 | Hermes 4 - Llama-3.1 405B (Non-reasoning) | 17.6 | 18.1 | $1.5 |
| 225 | Gemini 2.5 Flash-Lite (Reasoning) | 17.6 | 9.5 | $0.175 |
| 226 | Llama 3.1 Instruct 405B | 17.4 | 14.5 | $3.688 |
| 227 | GPT-4o (Nov '24) | 17.3 | 16.7 | $4.375 |
| 228 | DeepSeek R1 Distill Qwen 32B | 17.2 | - | $0.27 |
| 229 | Qwen3 VL 32B Instruct | 17.2 | 15.6 | $1.225 |
| 230 | GLM-4.6V (Non-reasoning) | 17.1 | 11.1 | $0.45 |
| 231 | Qwen3 235B A22B (Non-reasoning) | 17 | 14 | $1.225 |
| 232 | Gemini 2.0 Flash (experimental) | 16.8 | - | $0 |
| 233 | Magistral Small 1 | 16.8 | 11.1 | $0 |
| 234 | EXAONE 4.0 32B (Reasoning) | 16.7 | 14 | $0 |
| 235 | Qwen3 VL 8B (Reasoning) | 16.7 | 9.8 | $0.66 |
| 236 | Nova 2.0 Omni (Non-reasoning) | 16.6 | 13.8 | $0.85 |
| 237 | DeepSeek V3 (Dec '24) | 16.5 | 16.4 | $0.625 |
| 238 | Qwen3 32B (Reasoning) | 16.5 | 13.8 | $2.625 |
| 239 | DeepSeek R1 0528 Qwen3 8B | 16.4 | 7.8 | $0 |
| 240 | Qwen3.5 2B (Reasoning) | 16.3 | 3.5 | $0 |
| 241 | Qwen2.5 Max | 16.3 | - | $2.8 |
| 242 | Qwen3 14B (Reasoning) | 16.2 | 13.1 | $1.313 |
| 243 | Nanbeige4.1-3B | 16.1 | 8.9 | $0 |
| 244 | Qwen3 VL 30B A3B Instruct | 16.1 | 14.3 | $0.35 |
| 245 | Ministral 3 14B | 16 | 10.9 | $0.2 |
| 246 | DeepSeek R1 Distill Llama 70B | 16 | 11.4 | $0.875 |
| 247 | Hermes 4 - Llama-3.1 70B (Reasoning) | 16 | 14.4 | $0.198 |
| 248 | Gemini 1.5 Pro (Sep '24) | 16 | 23.6 | $0 |
| 249 | Solar Pro 2 (Preview) (Non-reasoning) | 16 | - | $0 |
| 250 | Claude 3.5 Sonnet (Oct '24) | 15.9 | 30.2 | $6 |
| 251 | Falcon-H1R-7B | 15.8 | 9.8 | $0 |
| 252 | DeepSeek R1 Distill Qwen 14B | 15.8 | - | $0 |
| 253 | Ling-flash-2.0 | 15.7 | 16.7 | $0.247 |
| 254 | Qwen3 Omni 30B A3B (Reasoning) | 15.6 | 12.7 | $0.43 |
| 255 | Qwen2.5 Instruct 72B | 15.6 | 11.9 | $0 |
| 256 | Sonar | 15.5 | - | $0 |
| 257 | Step3 VL 10B | 15.4 | 13.9 | $0 |
| 258 | Qwen3 30B A3B (Reasoning) | 15.3 | 11 | $0.75 |
| 259 | Devstral Small (Jul '25) | 15.2 | 12.1 | $0.15 |
| 260 | Sonar Pro | 15.2 | - | $0 |
| 261 | QwQ 32B-Preview | 15.2 | - | $0.135 |
| 262 | Mistral Large 2 (Nov '24) | 15.1 | 13.8 | $3 |
| 263 | Mistral Small 3.2 | 15.1 | 13.3 | $0.15 |
| 264 | GLM-4.5V (Reasoning) | 15.1 | 10.9 | $0.9 |
| 265 | Llama 3.1 Nemotron Ultra 253B v1 (Reasoning) | 15 | 13.1 | $0.9 |
| 266 | ERNIE 4.5 300B A47B | 15 | 14.5 | $0.485 |
| 267 | Qwen3 30B A3B 2507 Instruct | 15 | 14.2 | $0.35 |
| 268 | Solar Pro 2 (Reasoning) | 14.9 | 12.1 | $0 |
| 269 | NVIDIA Nemotron Nano 12B v2 VL (Reasoning) | 14.9 | 11.8 | $0.3 |
| 270 | Ministral 3 8B | 14.8 | 10 | $0.15 |
| 271 | NVIDIA Nemotron Nano 9B V2 (Reasoning) | 14.8 | 8.3 | $0.07 |
| 272 | NVIDIA Nemotron 3 Nano 4B | 14.7 | 10 | $0 |
| 273 | Qwen3.5 2B (Non-reasoning) | 14.7 | 4.9 | $0 |
| 274 | Gemini 2.0 Flash-Lite (Feb '25) | 14.7 | - | $0 |
| 275 | Llama Nemotron Super 49B v1.5 (Non-reasoning) | 14.6 | 10.5 | $0.175 |
| 276 | Llama 3.3 Instruct 70B | 14.5 | 10.7 | $0.64 |
| 277 | GPT-4o (May '24) | 14.5 | 24.2 | $7.5 |
| 278 | Gemini 2.0 Flash-Lite (Preview) | 14.5 | - | $0 |
| 279 | Mistral Small 3.1 | 14.5 | 13.9 | $0.15 |
| 280 | Qwen3 32B (Non-reasoning) | 14.5 | - | $1.225 |
| 281 | Llama 3.1 Nemotron Nano 4B v1.1 (Reasoning) | 14.4 | - | $0 |
| 282 | Kimi Linear 48B A3B Instruct | 14.4 | 14.2 | $0 |
| 283 | K2-V2 (low) | 14.4 | 10.5 | $0 |
| 284 | Llama 3.3 Nemotron Super 49B v1 (Non-reasoning) | 14.3 | 7.6 | $0 |
| 285 | Qwen3 VL 8B Instruct | 14.3 | 7.3 | $0.31 |
| 286 | Claude 3.5 Sonnet (June '24) | 14.2 | 26 | $6 |
| 287 | Qwen3 4B (Reasoning) | 14.2 | - | $0.398 |
| 288 | GPT-4o (ChatGPT) | 14.1 | - | $0 |
| 289 | Llama 3.1 Tulu3 405B | 14.1 | - | $0 |
| 290 | Ring-flash-2.0 | 14 | 10.6 | $0.247 |
| 291 | Pixtral Large | 14 | - | $3 |
| 292 | Olmo 3.1 32B Think | 13.9 | 9.8 | $0 |
| 293 | Grok 2 (Dec '24) | 13.9 | - | $0 |
| 294 | GPT-5 nano (minimal) | 13.8 | 14.2 | $0.138 |
| 295 | Gemini 1.5 Flash (Sep '24) | 13.8 | - | $0 |
| 296 | GPT-4 Turbo | 13.7 | 21.5 | $15 |
| 297 | Qwen3 VL 4B (Reasoning) | 13.7 | 6.7 | $0 |
| 298 | Solar Pro 2 (Non-reasoning) | 13.6 | 11.3 | $0 |
| 299 | Llama 4 Scout | 13.5 | 6.7 | $0.292 |
| 300 | Command A | 13.5 | 9.9 | $4.375 |
| 301 | Nova Pro | 13.5 | 11 | $1.4 |
| 302 | Llama 3.1 Nemotron Instruct 70B | 13.4 | 10.8 | $1.2 |
| 303 | Grok Beta | 13.3 | - | $0 |
| 304 | NVIDIA Nemotron Nano 9B V2 (Non-reasoning) | 13.2 | 7.5 | $0.086 |
| 305 | NVIDIA Nemotron 3 Nano 30B A3B (Non-reasoning) | 13.2 | 15.8 | $0.087 |
| 306 | Qwen3 8B (Reasoning) | 13.2 | 9 | $0.66 |
| 307 | Qwen2.5 Instruct 32B | 13.2 | - | $0 |
| 308 | GPT-4.1 nano | 13 | 11.2 | $0.175 |
| 309 | Mistral Large 2 (Jul '24) | 13 | - | $3 |
| 310 | Qwen2.5 Coder Instruct 32B | 12.9 | - | $0 |
| 311 | Qwen3 4B 2507 Instruct | 12.9 | 9.1 | $0 |
| 312 | GPT-4 | 12.8 | 13.1 | $37.5 |
| 313 | Qwen3 14B (Non-reasoning) | 12.8 | 12.4 | $0.613 |
| 314 | Gemini 2.5 Flash-Lite (Non-reasoning) | 12.7 | 7.4 | $0.175 |
| 315 | Mistral Small 3 | 12.7 | - | $0.15 |
| 316 | Nova Lite | 12.7 | 5.1 | $0.105 |
| 317 | GLM-4.5V (Non-reasoning) | 12.7 | 10.8 | $0.9 |
| 318 | Hermes 4 - Llama-3.1 70B (Non-reasoning) | 12.6 | 9.2 | $0.198 |
| 319 | GPT-4o mini | 12.6 | - | $0.263 |
| 320 | Llama 3.1 Instruct 70B | 12.5 | 10.9 | $0.56 |
| 321 | DeepSeek-V2.5 (Dec '24) | 12.5 | - | $0 |
| 322 | Qwen3 4B (Non-reasoning) | 12.5 | - | $0.188 |
| 323 | Qwen3 30B A3B (Non-reasoning) | 12.5 | 13.3 | $0.35 |
| 324 | Sarvam 30B (high) | 12.3 | 7.9 | $0 |
| 325 | Gemini 2.0 Flash Thinking Experimental (Dec '24) | 12.3 | - | $0 |
| 326 | Claude 3 Haiku | 12.3 | 6.7 | $0.5 |
| 327 | DeepSeek-V2.5 | 12.3 | - | $0 |
| 328 | Olmo 3.1 32B Instruct | 12.2 | 5.6 | $0.3 |
| 329 | Mistral Saba | 12.1 | - | $0 |
| 330 | DeepSeek R1 Distill Llama 8B | 12.1 | - | $0 |
| 331 | Olmo 3 32B Think | 12.1 | 10.5 | $0 |
| 332 | R1 1776 | 12 | - | $0 |
| 333 | Gemini 1.5 Pro (May '24) | 12 | 19.8 | $0 |
| 334 | Reka Flash (Sep '24) | 12 | - | $0.35 |
| 335 | Qwen2.5 Turbo | 12 | - | $0.087 |
| 336 | Llama 3.2 Instruct 90B (Vision) | 11.9 | - | $0.72 |
| 337 | Solar Mini | 11.9 | - | $0.15 |
| 338 | Llama 3.1 Instruct 8B | 11.8 | 4.9 | $0.1 |
| 339 | Grok-1 | 11.7 | - | $0 |
| 340 | EXAONE 4.0 32B (Non-reasoning) | 11.7 | 9.4 | $0 |
| 341 | Qwen2 Instruct 72B | 11.7 | - | $0 |
| 342 | Ministral 3 3B | 11.2 | 4.8 | $0.1 |
| 343 | Gemini 1.5 Flash-8B | 11.1 | - | $0 |
| 344 | DeepHermes 3 - Mistral 24B Preview (Non-reasoning) | 10.9 | - | $0 |
| 345 | Jamba 1.7 Large | 10.9 | 7.8 | $3.5 |
| 346 | Granite 4.0 H Small | 10.8 | 8.5 | $0.107 |
| 347 | Qwen3 Omni 30B A3B Instruct | 10.7 | 7.2 | $0.43 |
| 348 | Jamba 1.5 Large | 10.7 | - | $3.5 |
| 349 | DeepSeek-Coder-V2 | 10.6 | - | $0 |
| 350 | OLMo 2 32B | 10.6 | 2.7 | $0 |
| 351 | Hermes 3 - Llama-3.1 70B | 10.6 | - | $0.3 |
| 352 | Jamba 1.6 Large | 10.6 | - | $3.5 |
| 353 | Qwen3 8B (Non-reasoning) | 10.6 | 7.1 | $0.31 |
| 354 | LFM2 24B A2B | 10.5 | 3.6 | $0.052 |
| 355 | Qwen3.5 0.8B (Reasoning) | 10.5 | 0 | $0 |
| 356 | Gemini 1.5 Flash (May '24) | 10.5 | - | $0 |
| 357 | Phi-4 | 10.4 | 11.2 | $0.219 |
| 358 | Gemma 3 27B Instruct | 10.3 | 9.6 | $0 |
| 359 | Nova Micro | 10.3 | 4.1 | $0.061 |
| 360 | Claude 3 Sonnet | 10.3 | - | $6 |
| 361 | Mistral Small (Sep '24) | 10.2 | - | $0.3 |
| 362 | NVIDIA Nemotron Nano 12B v2 VL (Non-reasoning) | 10.1 | 5.9 | $0.3 |
| 363 | Gemma 3n E4B Instruct Preview (May '25) | 10.1 | - | $0 |
| 364 | Gemini 1.0 Ultra | 10.1 | 17.6 | $0 |
| 365 | Phi-3 Mini Instruct 3.8B | 10.1 | 3 | $0 |
| 366 | Phi-4 Multimodal Instruct | 10 | - | $0 |
| 367 | Qwen2.5 Coder Instruct 7B | 10 | - | $0 |
| 368 | Qwen3.5 0.8B (Non-reasoning) | 9.9 | 1 | $0 |
| 369 | Mistral Large (Feb '24) | 9.9 | - | $6 |
| 370 | Mixtral 8x22B Instruct | 9.8 | - | $0 |
| 371 | Llama 3.2 Instruct 3B | 9.7 | - | $0.085 |
| 372 | Llama 2 Chat 7B | 9.7 | - | $0.1 |
| 373 | Jamba Reasoning 3B | 9.6 | 2.5 | $0 |
| 374 | Qwen3 VL 4B Instruct | 9.6 | 4.5 | $0 |
| 375 | Reka Flash 3 | 9.5 | 8.9 | $0.35 |
| 376 | Qwen1.5 Chat 110B | 9.5 | - | $0 |
| 377 | Olmo 3 7B Think | 9.4 | 7.6 | $0 |
| 378 | Claude 2.1 | 9.3 | 14 | $0 |
| 379 | OLMo 2 7B | 9.3 | 1.2 | $0 |
| 380 | Molmo 7B-D | 9.2 | 1.2 | $0 |
| 381 | Ling-mini-2.0 | 9.2 | 5 | $0 |
| 382 | Claude 2.0 | 9.1 | 12.9 | $0 |
| 383 | DeepSeek R1 Distill Qwen 1.5B | 9.1 | - | $0 |
| 384 | DeepSeek-V2-Chat | 9.1 | - | $0 |
| 385 | GPT-3.5 Turbo | 9 | 10.7 | $0.75 |
| 386 | Mistral Small (Feb '24) | 9 | - | $1.5 |
| 387 | Mistral Medium | 9 | - | $4.088 |
| 388 | Llama 3 Instruct 70B | 8.9 | 6.8 | $0.871 |
| 389 | Gemma 3 12B Instruct | 8.8 | 6.3 | $0 |
| 390 | LFM 40B | 8.8 | - | $0 |
| 391 | Arctic Instruct | 8.8 | - | $0 |
| 392 | Qwen Chat 72B | 8.8 | - | $0 |
| 393 | Llama 3.2 Instruct 11B (Vision) | 8.7 | 4.3 | $0.16 |
| 394 | PALM-2 | 8.6 | 4.6 | $0 |
| 395 | Gemini 1.0 Pro | 8.5 | - | $0 |
| 396 | DeepSeek Coder V2 Lite Instruct | 8.5 | - | $0 |
| 397 | Phi-4 Mini Instruct | 8.4 | 3.6 | $0 |
| 398 | Llama 2 Chat 70B | 8.4 | - | $0 |
| 399 | Llama 2 Chat 13B | 8.4 | - | $0 |
| 400 | DeepSeek LLM 67B Chat (V1) | 8.4 | - | $0 |
| 401 | Sarvam M (Reasoning) | 8.4 | 7.5 | $0 |
| 402 | Exaone 4.0 1.2B (Reasoning) | 8.3 | 3.1 | $0 |
| 403 | OpenChat 3.5 (1210) | 8.3 | - | $0 |
| 404 | DBRX Instruct | 8.3 | - | $0 |
| 405 | Command-R+ (Apr '24) | 8.3 | - | $6 |
| 406 | Olmo 3 7B Instruct | 8.2 | 3.4 | $0.125 |
| 407 | LFM2.5-1.2B-Thinking | 8.1 | 1.4 | $0 |
| 408 | Exaone 4.0 1.2B (Non-reasoning) | 8.1 | 2.5 | $0 |
| 409 | Jamba 1.7 Mini | 8.1 | 3.1 | $0 |
| 410 | LFM2.5-1.2B-Instruct | 8 | 0.8 | $0 |
| 411 | LFM2 2.6B | 8 | 1.4 | $0 |
| 412 | Granite 4.0 H 1B | 8 | 2.7 | $0 |
| 413 | Jamba 1.5 Mini | 8 | - | $0.25 |
| 414 | Qwen3 1.7B (Reasoning) | 8 | 1.4 | $0.398 |
| 415 | Jamba 1.6 Mini | 7.9 | - | $0.25 |
| 416 | Gemma 3 270M | 7.7 | 0 | $0 |
| 417 | Granite 4.0 Micro | 7.7 | 5 | $0 |
| 418 | Apertus 70B Instruct | 7.7 | 1.9 | $1.345 |
| 419 | Mixtral 8x7B Instruct | 7.7 | - | $0.54 |
| 420 | DeepHermes 3 - Llama-3.1 8B Preview (Non-reasoning) | 7.6 | - | $0 |
| 421 | Llama 65B | 7.4 | - | $0 |
| 422 | Qwen Chat 14B | 7.4 | - | $0 |
| 423 | Claude Instant | 7.4 | 7.8 | $0 |
| 424 | Mistral 7B Instruct | 7.4 | - | $0.25 |
| 425 | Command-R (Mar '24) | 7.4 | - | $0.75 |
| 426 | Molmo2-8B | 7.3 | 4.4 | $0 |
| 427 | Granite 4.0 1B | 7.3 | 2.9 | $0 |
| 428 | LFM2 8B A1B | 7 | 2.3 | $0 |
| 429 | Granite 3.3 8B (Non-reasoning) | 7 | 3.4 | $0.085 |
| 430 | Qwen3 1.7B (Non-reasoning) | 6.8 | 2.3 | $0.188 |
| 431 | Qwen3 0.6B (Reasoning) | 6.5 | 0.9 | $0.398 |
| 432 | Gemma 3n E4B Instruct | 6.4 | 4.2 | $0.025 |
| 433 | Llama 3 Instruct 8B | 6.4 | 4 | $0.07 |
| 434 | Gemma 3 4B Instruct | 6.3 | 2.9 | $0 |
| 435 | Llama 3.2 Instruct 1B | 6.3 | 0.6 | $0.1 |
| 436 | LFM2 1.2B | 6.3 | 0.8 | $0 |
| 437 | LFM2.5-VL-1.6B | 6.2 | 1 | $0 |
| 438 | Granite 4.0 350M | 6.1 | 0.3 | $0 |
| 439 | Apertus 8B Instruct | 5.9 | 1.4 | $0.125 |
| 440 | Qwen3 0.6B (Non-reasoning) | 5.7 | 1.4 | $0.188 |
| 441 | Gemma 3 1B Instruct | 5.5 | 0.2 | $0 |
| 442 | Granite 4.0 H 350M | 5.4 | 0.6 | $0 |
| 443 | Gemma 3n E2B Instruct | 4.8 | 2.2 | $0 |
| 444 | Tiny Aya Global | 4.7 | 1.2 | $0 |
| 445 | GPT-5.4 Pro (xhigh) | - | - | $67.5 |
| 446 | Gemini 3 Deep Think | - | - | $0 |
| 447 | Cogito v2.1 (Reasoning) | - | 24.8 | $1.25 |
| 448 | Mi:dm K 2.5 Pro Preview | - | 11.9 | $0 |
| 449 | GPT-4o Realtime (Dec '24) | - | - | $0 |
| 450 | GPT-4o mini Realtime (Dec '24) | - | - | $0 |
| 451 | GPT-3.5 Turbo (0613) | - | - | $0 |
榜单解读建议
参考 AI 大模型排行榜 时，应综合考虑“综合指数”与“成本价格”。如果您是开发者，编程能力 (Coding) 是更核心的指标。
值品工具箱同步的 AI 大模型排行榜 数据每 24 小时更新，确保您获取到最新的模型性能对比。
指标说明
- ● 综合指数：评估通用理解与逻辑。
- ● 价格 $/1M：混合 3:1 输入输出比的平均成本。
- ● 编程能力：衡量代码生成的准确性。
AI 大模型排行榜 常见问题 (FAQ)
Q1: AI 大模型排行榜 的数据多久更新？
AI 大模型排行榜 数据每 24 小时自动抓取一次，确保最新模型加入列表。
Q2: 这个 AI 大模型排行榜 包含国产模型吗？
是的，只要国产模型通过了 Artificial Analysis 的全球测评，就会出现在 AI 大模型排行榜 中。
Q3: 综合指数在 AI 大模型排行榜 中代表什么？
它代表模型的全能表现。AI 大模型排行榜 通过加权算法给出这个综合评分。
Q4: 如何在 AI 大模型排行榜 中查找性价比最高的游戏？
在 AI 大模型排行榜 页面中，您可以点击“价格”标题进行排序，寻找低价高分的模型。
Q5: AI 大模型排行榜 的编程能力测试准吗？
AI 大模型排行榜 参考了 LiveCodeBench 等权威基准测试，具有极高的参考价值。
Q6: 为什么有的新模型没进入 AI 大模型排行榜？
模型进入 AI 大模型排行榜 需要经过一系列测试，通常在新模型发布后数日内会完成更新。
Q7: AI 大模型排行榜 中的价格计算标准是什么？
价格是基于百万 Token 的调用成本，由 AI 大模型排行榜 统一混合计算得出。
Q8: 手机上能查看 AI 大模型排行榜 吗？
当然可以。AI 大模型排行榜 进行了移动端响应式深度优化。
Q9: AI 大模型排行榜 这个工具免费吗？
是的，由值品工具箱免费提供 AI 大模型排行榜 信息查询服务。
Q10: 我该怎么利用 AI 大模型排行榜 做选型？
如果您需要智能客服，参考 AI 大模型排行榜 的综合指数；如果做翻译，参考编程外的语言指标。