LOADING加载中
42 个模型
| 排名 | 模型 | 评分 | 评分下界 | 评分上界 | 方差 | 投票数 | 分类 | 源数据更新 |
|---|---|---|---|---|---|---|---|---|
| #1 | Qwen3.8-Maxalibaba · Proprietary | 1,246.19 | 1,273.38 | 48.14 | 1,998 | overall | 2026-11-09 | |
| #2 | GPT-5.6 Sol (xHigh)openai · Proprietary | 1,251.39 | 1,264.29 | 10.82 | 62,364 | overall | 2026-11-09 | |
| #3 | Claude Opus 4.6 (Search)anthropic · Proprietary | 1,247.03 | 1,256.32 | 5.61 | 166,732 | overall | 2026-11-09 | |
| #4 | GPT-5.5 (Search)openai · Proprietary | 1,240.69 | 1,251.03 | 6.96 | 121,489 | overall | 2026-11-09 | |
| #5 | Claude Opus 5 (High)anthropic · Proprietary | 1,232.23 | 1,245.82 | 12.03 | 34,000 | overall | 2026-11-09 | |
| #6 | Claude Fable 5.1 (High)anthropic · Proprietary | 1,230.41 | 1,244.2 | 12.37 | 33,769 | overall | 2026-11-09 | |
| #7 | GPT-6 Astra (Max)openai · Proprietary | 1,222.94 | 1,238.68 | 16.13 | 17,456 | overall | 2026-11-09 | |
| #8 | GPT-6.1 Sol (Max)openai · Proprietary | 1,222.57 | 1,238.16 | 15.82 | 26,556 | overall | 2026-11-09 | |
| #9 | Claude Opus 4.7anthropic · Proprietary | 1,225.11 | 1,235.09 | 6.47 | 121,890 | overall | 2026-11-09 | |
| #10 | Claude Fable 5 (High)anthropic · Proprietary | 1,221.51 | 1,234.8 | 11.49 | 69,528 | overall | 2026-11-09 | |
| #11 | ERNIE 5.1baidu · Proprietary | 1,217.21 | 1,236.39 | 23.93 | 3,925 | overall | 2026-11-09 | |
| #12 | Claude Sonnet 4.6 (Search)anthropic · Proprietary | 1,215.99 | 1,225.25 | 5.58 | 167,499 | overall | 2026-11-09 | |
| #13 | Grok 4.5xai · Proprietary | 1,210.07 | 1,221.57 | 8.6 | 105,282 | overall | 2026-11-09 | |
| #14 | Claude Opus 5.5 (Medium)anthropic · Proprietary | 1,206.55 | 1,220.11 | 11.97 | 31,730 | overall | 2026-11-09 | |
| #15 | Gemini 3.1 Pro (Grounding)google · Proprietary | 1,205.78 | 1,215.02 | 5.55 | 171,494 | overall | 2026-11-09 | |
| #16 | GPT-5.2 (Search)openai · Proprietary | 1,201.45 | 1,212.7 | 8.23 | 54,658 | overall | 2026-11-09 | |
| #17 | Gemini 3 Pro (Grounding)google · Proprietary | 1,201.02 | 1,212.12 | 8.02 | 38,324 | overall | 2026-11-09 | |
| #18 | Grok 4.20 Multi-Agent Beta (0309)xai · Proprietary | 1,199.61 | 1,210.7 | 8.01 | 121,069 | overall | 2026-11-09 | |
| #19 | Claude Opus 4.8anthropic · Proprietary | 1,196.65 | 1,208.18 | 8.64 | 101,197 | overall | 2026-11-09 | |
| #20 | GPT-5.1 (Search)openai · Proprietary | 1,194.15 | 1,204.68 | 7.22 | 61,989 | overall | 2026-11-09 | |
| #21 | GPT-5.4 (Search)openai · Proprietary | 1,192.47 | 1,203.59 | 8.05 | 121,580 | overall | 2026-11-09 | |
| #22 | Gemini 3 Flash (Grounding)google · Proprietary | 1,191.16 | 1,199.99 | 5.07 | 189,590 | overall | 2026-11-09 | |
| #23 | Grok 4.20 Beta 1xai · Proprietary | 1,184.84 | 1,196.93 | 9.52 | 55,902 | overall | 2026-11-09 | |
| #24 | Claude Sonnet 5 (Search)anthropic · Proprietary | 1,181.31 | 1,194.18 | 10.79 | 71,297 | overall | 2026-11-09 | |
| #25 | Claude Opus 4.5 (Search)anthropic · Proprietary | 1,173.51 | 1,184.84 | 8.36 | 63,904 | overall | 2026-11-09 | |
| #26 | GPT-6 Sol (Max)openai · Proprietary | 1,167.02 | 1,181.92 | 14.46 | 13,499 | overall | 2026-11-09 | |
| #27 | GPT-5.2 (Search, Non-Reasoning)openai · Proprietary | 1,166.28 | 1,177.46 | 8.15 | 78,443 | overall | 2026-11-09 | |
| #28 | Grok 4.1 Fast (Search)xai · Proprietary | 1,166.62 | 1,176.64 | 6.53 | 84,455 | overall | 2026-11-09 | |
| #29 | Grok 4 Fast (Search)xai · Proprietary | 1,165.91 | 1,175.35 | 5.8 | 42,640 | overall | 2026-11-09 | |
| #30 | Grok 4.3xai · Proprietary | 1,161.59 | 1,171.27 | 6.1 | 122,703 | overall | 2026-11-09 | |
| #31 | GPT-6 Luna (Max)openai · Proprietary | 1,153.23 | 1,167.1 | 12.51 | 46,776 | overall | 2026-11-09 | |
| #32 | Claude Sonnet 4.5 (Search)anthropic · Proprietary | 1,154.01 | 1,164.37 | 6.98 | 139,468 | overall | 2026-11-09 | |
| #33 | Claude Opus 4.1 (Search)anthropic · Proprietary | 1,143.25 | 1,153.14 | 6.36 | 78,777 | overall | 2026-11-09 | |
| #34 | o3 (Search)openai · Proprietary | 1,138.85 | 1,150.21 | 8.39 | 20,655 | overall | 2026-11-09 | |
| #35 | Gemini 2.5 Pro (Grounding)google · Proprietary | 1,138.63 | 1,148.16 | 5.91 | 85,505 | overall | 2026-11-09 | |
| #36 | Grok 4 (Search)xai · Proprietary | 1,135.72 | 1,148.05 | 9.89 | 19,069 | overall | 2026-11-09 | |
| #37 | Sonar Reasoning Pro (High)perplexity · Proprietary | 1,132.91 | 1,144.72 | 9.07 | 29,197 | overall | 2026-11-09 | |
| #38 | GPT-5 (Search)openai · Proprietary | 1,127.03 | 1,139.15 | 9.55 | 20,742 | overall | 2026-11-09 | |
| #39 | Sonar Pro (High)perplexity · Proprietary | 1,123.8 | 1,135.59 | 9.06 | 28,684 | overall | 2026-11-09 | |
| #40 | Claude Opus 4 (Search)anthropic · Proprietary | 1,121.55 | 1,132.54 | 7.86 | 31,275 | overall | 2026-11-09 | |
| #41 | Diffbot-Small-XLdiffbot · Apache 2.0 | 1,013.47 | 1,030.61 | 19.12 | 6,339 | overall | 2026-11-09 | |
| #42 | GPT-4o (Search)openai · Proprietary | 995.1 | 1,017.63 | 33.05 | 3,392 | overall | 2026-11-09 |
用113个从零编写的长程真实软件工程任务测评编程Agent，同时刻画准确率(Pass@1)与真实花费/步数的性价比曲线，避免题目训练污染。
在新标签页打开独立跑分，不依赖厂商自报数据；综合 Intelligence Index 覆盖推理/编程/科学/知识，并联动速度、价格、上下文窗口等工程指标。
在新标签页打开每月更新全新题目防污染，全自动客观判分（唯一正确答案），聚焦数学/代码/推理/数据分析，学术圈公认权威。
在新标签页打开中文原创题库，深度贴合中文语境与政策文化，覆盖数学推理、代码、幻觉控制、智能体规划六大能力。
在新标签页打开上海AI实验室出品，20余家企业共建，五大能力维度、70+ 数据集，中英文模型均覆盖，并含真人投票的 Compass Arena。
在新标签页打开本页对战数据的官方原始来源，累计投票超过700万次，覆盖370+模型，逐票更新，界面本身也值得参考。
在新标签页打开
