大模型排行榜
Arena比较模型针对图像给出的回答,由参与者投票。
Arena AI 原榜 · 榜单日期
| 排名 | 模型 | 评分 | 对战票数 |
|---|---|---|---|
| 1 | Claude Fable 5 (High)anthropic | 1,3251318–1332 | 15,449 |
| 2 | Claude Fable 5.1 (Max)anthropic | 1,3221314–1330 | 7,185 |
| 3 | Claude Opus 5.5 (High)anthropic | 1,3221313–1331 | 5,903 |
| 4 | Claude Opus 4.7anthropic | 1,3171311–1324 | 24,930 |
| 5 | Claude Opus 5 (High)anthropic | 1,3171311–1324 | 19,229 |
| 6 | Gemini 3.8 Flash (High)google | 1,3161309–1324 | 9,782 |
| 7 | Claude Opus 4.6 (High)anthropic | 1,3151309–1321 | 22,067 |
| 8 | Claude Opus 4.7 (High)anthropic | 1,3141307–1320 | 22,530 |
| 9 | Gemini 3.7 Flash (High)google | 1,3141305–1322 | 6,776 |
| 10 | Claude Fable 5.2 (Max)anthropic | 1,3131295–1331 | 1,176 |
| 11 | Claude Opus 4.6anthropic | 1,3121306–1318 | 26,552 |
| 12 | Gemini 3.5 Flash (High)google | 1,3081301–1314 | 15,228 |
| 13 | Qwen3.8-Maxalibaba | 1,3071299–1314 | 10,833 |
| 14 | Muse Spark 1.3 (Max)meta | 1,3061299–1314 | 9,287 |
| 15 | Gemini 3.5 Flash (Medium)google | 1,3061300–1313 | 15,320 |
| 16 | Muse Sparkmeta | 1,3061297–1315 | 5,835 |
| 17 | Gemini 3 Progoogle | 1,3051297–1312 | 13,482 |
| 18 | Muse Spark 1.2 (xHigh)meta | 1,3041290–1319 | 1,922 |
| 19 | GPT-5.4 (High)openai | 1,3011295–1306 | 31,907 |
| 20 | Muse Spark 1.3 (xHigh)meta | 1,3001293–1308 | 9,809 |
| 21 | Gemini 3.6 Flash (High)google | 1,2971289–1304 | 10,275 |
| 22 | Gemini 3.1 Pro Previewgoogle | 1,2961292–1301 | 48,038 |
| 23 | GLM-5.3-Flashzai | 1,2961291–1302 | 36,497 |
| 24 | GPT-5.5openai | 1,2961290–1302 | 29,925 |
| 25 | GPT-5.5 (High)openai | 1,2951290–1301 | 28,221 |
| 26 | Claude Opus 4.8 (High)anthropic | 1,2941288–1301 | 18,113 |
| 27 | GPT-5.4openai | 1,2911285–1298 | 22,546 |
| 28 | Muse Spark 1.1meta | 1,2911284–1298 | 13,311 |
| 29 | Claude Opus 4.8anthropic | 1,2881282–1294 | 18,588 |
| 30 | Grok 4.5xai | 1,2881281–1295 | 13,903 |
| 31 | Claude Sonnet 5.5 (xHigh)anthropic | 1,2861277–1295 | 5,388 |
| 32 | Gemini 3 Flashgoogle | 1,2851280–1290 | 38,733 |
| 33 | GPT-6.1 Sol (Max)openai | 1,2841274–1293 | 5,043 |
| 34 | Claude Sonnet 4.6anthropic | 1,2821276–1289 | 27,131 |
| 35 | Kimi K2.6moonshot | 1,2821276–1289 | 16,241 |
| 36 | GPT-5.6 Sol (xHigh)openai | 1,2821275–1289 | 12,244 |
| 37 | Gemma 4 31Bgoogle | 1,2791274–1285 | 42,217 |
| 38 | DeepSeek-V4.1-Flash-20260910 (Max)deepseek | 1,2781268–1288 | 4,192 |
| 39 | Qwen3.7-Plusalibaba | 1,2771271–1284 | 15,084 |
| 40 | GPT-6 Astra (Max)openai | 1,2761268–1285 | 6,436 |
| 41 | Dola Seed 2.0 Probytedance | 1,2741267–1281 | 10,914 |
| 42 | Claude Sonnet 5 (High)anthropic | 1,2721265–1278 | 14,701 |
| 43 | Qwen3.8-27Balibaba | 1,2711264–1278 | 10,910 |
| 44 | GPT-5.6 Terra (xHigh)openai | 1,2701263–1277 | 13,166 |
| 45 | Kimi K2.5 (Thinking)moonshot | 1,2691264–1275 | 27,947 |
| 46 | Gemini 3.5 Flash-Litegoogle | 1,2681261–1275 | 10,713 |
| 47 | GPT-5.2 Chat (20260210)openai | 1,2681261–1274 | 16,336 |
| 48 | MiMo-V2.6-Proxiaomi | 1,2661251–1281 | 1,812 |
| 49 | Gemini 3 Flash (Minimal)google | 1,2661260–1271 | 36,810 |
| 50 | Grok 4.6 (High)xai | 1,2651257–1273 | 8,578 |