← 全部基准

智能体竞技场

来自 Wikiprompt,自由的提示词百科

社区 Elo 排行榜,每日从 LMArena 对战汇总。先是总榜,然后是分类明细。

最后更新: 2026-10-12 · 52 个模型

排名模型机构Elo 分数
1Claude Opus 5.5 (High)anthropic
2GPT 6 Astra (Max)openai
3Claude Fable 5.1 (Max)anthropic
4Claude Sonnet 5.5 (Max)anthropic
5GPT 6.1 Sol (Max)openai
6GPT 6 Sol (Max)openai
7Gemini 4 Argon (High)google
8Claude Fable 5 (High)anthropic
9Claude Opus 5 (Max)anthropic
10Claude Opus 5 (High)anthropic
11Claude Opus 4.8 (High)anthropic
12GPT 5.6 Sol (xHigh)openai
13Claude Sonnet 5 (High)anthropic
14GPT 5.5 (xHigh)openai
15Muse Spark 1.3 (Max)meta
16Deepseek V4.1 Flash (Max)deepseek
17Kimi K3 (Max)moonshot
18Hy4 previewtencent
19MiMo V2.6 Proxiaomi
20GLM 5.2 (Max)zai
21Grok 4.7 (xHigh)xai
22Gemini 3.8 Flash (High)google
23Qwen3.8 Maxalibaba
24GLM 5.3 (Max)zai
25GPT 6 Luna (Max)openai
26DeepSeek V4 Pro (High) (0813)deepseek
27Grok 4.5xai
28GPT 5.4 (High)openai
29Grok 4.6 (xHigh)xai
30GPT 5.5openai
31GPT 5.6 Terra (xHigh)openai
32MiMo V2.6 Flashxiaomi
33Step 5 Previewstepfun
34Qwen3.8 Flash Nextalibaba
35GLM 5.3 Flashzai
36GPT 5.6 Luna (xHigh)openai
37Qwen 3.8 27Balibaba
38Gemini 3.7 Flash (High)google
39Muse Spark 1.2 (xHigh)meta
40Muse Spark 1.1meta
41Hy3tencent
42Qwen3.7 Maxalibaba
43mistral-large-4mistral
44Minimax M3minimax
45Qwen3.7 Plusalibaba
46Mimo V2.5 Proxiaomi
47Gemini 3.6 Flash (High)google
48Gemini 3.1 Pro Previewgoogle
49Inkling Smallthinky
50Inklingthinky
51Mistral Medium 3.5mistral
52Solar Pro 4upstage

参见

数据来源与方法

排行榜汇总自开放的社区评测。Elo 分数、置信区间与投票数来自 LMArena(社区投票的模型对战,以开放数据集发布)。价格与上下文元数据来自 OpenRouter 公开 API。Wikiprompt 每日快照并保留历史,不进行自有评测。

  1. LMArena leaderboard dataset (Hugging Face)
  2. LMArena leaderboard