Design Arena · 网站设计
定期获取并展示公开的基准测试资料(LMArena 用户投票排名、Epoch AI 测试分数、OpenRouter 使用量排名)。每个榜单均标注数据发布日期和获取日期。这些分数并非本站自行评定。
如何解读此排行榜
衡量什么: 这是 Design Arena 中,人们比较根据相同请求生成的网站并投票得出的排名。
如何解读分数: 采用 Elo 方式计算相对得分,得分越高,表示人们更经常偏好该模型的结果。同时显示胜率。
注意事项: 这是反映设计偏好的投票,并不直接衡量代码质量或可维护性。
本次排名的解读要点
- 最高分为 Muse Spark 1.3 (xhigh) 的 1365。
- 前 10 个模型中,Meta 的模型有 3 个,数量最多。
网站设计 Design Arena
发布日期:2026-10-05 · 获取日期:2026-10-05
图表仅显示同一模型各推理设置(high、max 等)中排名最高的一项。表格列出所有设置各自的排名。
查看表格(前 60 名)
| 排名 | 模型 | 开发方 | 分数 | 胜率 |
|---|---|---|---|---|
| 1 | #1Muse Spark 1.3 (xhigh) | Meta | 1365 | 58.4% |
| 2 | #2Muse Spark 1.3 Max | Meta | 1352 | 56% |
| 3 | #3Kimi K3 | Moonshot AI | 1344 | 59.9% |
| 4 | #4Claude Opus 5.5 | Anthropic | 1343 | 57.7% |
| 5 | #5Claude Fable 5.1 | Anthropic | 1320 | 56.3% |
| 6 | #6Muse Spark 1.2 | Meta | 1320 | 52.2% |
| 7 | #7MiMo-V2.6-Pro | Xiaomi | 1319 | 51.6% |
| 8 | #8Claude Opus 5 | Anthropic | 1315 | 56.1% |
| 9 | #9Gemini 3.7 Flash | 1311 | 54.5% | |
| 10 | #10Gemini 3.8 Flash | 1309 | 50.4% | |
| 11 | #11ZGLM-5.3 | Z.ai | 1308 | 55% |
| 12 | #12Gemini 3.6 Flash | 1306 | 55.4% | |
| 13 | #13Claude Fable 5 | Anthropic | 1302 | 56% |
| 14 | #14Claude Opus 4.6 | Anthropic | 1298 | 61.2% |
| 15 | #15xGrok 4.6 | xAI | 1298 | 52.6% |
| 16 | #16ZGLM 5.2 | Z.ai | 1294 | 50.1% |
| 17 | #17Claude Opus 4.6 (Thinking) | Anthropic | 1293 | 60.2% |
| 18 | #18Claude Sonnet 4.6 | Anthropic | 1291 | 60% |
| 19 | #19ZGLM 5.1 | Z.ai | 1290 | 55.1% |
| 20 | #20xGrok 4.5 | xAI | 1289 | 52.9% |
| 21 | #21Claude Sonnet 5 | Anthropic | 1283 | 52.9% |
| 22 | #22Qwen3.8 Max | Alibaba | 1283 | 51.5% |
| 23 | #23ZGLM-5.3-Flash | Z.ai | 1280 | 48.5% |
| 24 | #24Qwen3.7 Max | Alibaba | 1280 | 54.8% |
| 25 | #25Qwen3.7 Plus | Alibaba | 1277 | 51% |
| 26 | #26Muse Spark 1.1 | Meta | 1276 | 51.6% |
| 27 | #27Kimi K2.6 | Moonshot AI | 1276 | 54.2% |
| 28 | #28MiMo-V2.5-Pro | Xiaomi | 1275 | 51.2% |
| 29 | #29ZGLM 5.1 | Z.ai | 1272 | 52.5% |
| 30 | #30Gemini 3.5 Flash | 1272 | 53.9% | |
| 31 | #31Kimi K2.7 Code | Moonshot AI | 1272 | 51.7% |
| 32 | #32MiMo-V2.5 | Xiaomi | 1272 | 53.8% |
| 33 | #33ZGLM 5 Turbo | Z.ai | 1270 | 53.4% |
| 34 | #34MiniMax M3 | MiniMax | 1264 | 51.9% |
| 35 | #35Claude Opus 4.8 | Anthropic | 1263 | 53.1% |
| 36 | #36OGPT-5.5 | OpenAI | 1263 | 52.8% |
| 37 | #37Gemini 3 Pro Preview | 1259 | 64.4% | |
| 38 | #38Kimi K2.5 (Thinking) | Moonshot AI | 1255 | 55.2% |
| 39 | #39ZGLM 5 | Z.ai | 1254 | 55% |
| 40 | #40Claude Opus 4.5 | Anthropic | 1253 | 59.7% |
| 41 | #41MiniMax M2.7 | MiniMax | 1249 | 51.5% |
| 42 | #42Qwen3.6 Plus | Alibaba | 1248 | 50.4% |
| 43 | #43DeepSeek-V4-Flash-0731 | DeepSeek | 1245 | 46.7% |
| 44 | #44DeepSeek-V4-Pro | DeepSeek | 1241 | 50.5% |
| 45 | #45Gemini 3.1 Pro Preview | 1240 | 53.8% | |
| 46 | #46xGrok 4.7 | xAI | 1239 | 43.2% |
| 47 | #47ZGLM 5V Turbo | Z.ai | 1237 | 49.2% |
| 48 | #48Claude 3.7 Sonnet | Anthropic | 1236 | 58.3% |
| 49 | #49xGrok 4.20 Beta (Reasoning) | xAI | 1236 | 51.8% |
| 50 | #50ZGLM 4.7 | Z.ai | 1232 | 55.3% |
| 51 | #51MiniMax M2.5 | MiniMax | 1229 | 57.5% |
| 52 | #52TInkling | Thinkingmachines | 1229 | 44.3% |
| 53 | #53NNex N2 Pro | Nex Agi | 1228 | 45.8% |
| 54 | #54xGrok 4.20 Beta | xAI | 1227 | 50.5% |
| 55 | #55OGPT-5.4 (Design Skill, Medium) | OpenAI | 1226 | 52.5% |
| 56 | #56OGPT-5.4 (Medium) | OpenAI | 1222 | 52.7% |
| 57 | #57DeepSeek-V4-Flash | DeepSeek | 1214 | 49.1% |
| 58 | #58MiMo-V2-Flash | Xiaomi | 1209 | 49.9% |
| 59 | #59MiniMax M2.1 | MiniMax | 1208 | 55.4% |
| 60 | #60OGPT-5.4 (Low) | OpenAI | 1205 | 50.5% |
Source: Design Arena (www.designarena.ai) via OpenRouter (openrouter.ai/rankings). Licensed under CC BY 4.0.
数据:LMArena leaderboard dataset(CC BY 4.0)。修改:各榜单仅显示前 50 名,分数四舍五入。 https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset公司标志为各公司的商标,仅用于帮助区分(图标:Simple Icons)。数据:Epoch AI — Capabilities & benchmarking(CC BY 4.0)。 https://epoch.ai/benchmarksSource: Artificial Analysis (artificialanalysis.ai) via OpenRouter (openrouter.ai/rankings). Source: OpenRouter evals (openrouter.ai) via OpenRouter (openrouter.ai/rankings). Source: Design Arena (www.designarena.ai) via OpenRouter (openrouter.ai/rankings). Licensed under CC BY 4.0. Source: OpenRouter (openrouter.ai/rankings), as of 2026-10-05. Licensed under CC BY 4.0. https://openrouter.ai/rankings