大模型排行榜
Arena评测同步于
评分说明与数据来源
综合榜采用 Epoch 能力指数(ECI)。各分类使用页面标注的单项评测,展示该评测的原始百分比分,不合并不同榜单的分数。
评测有多个运行记录时,优先取日期最近的一次。模型详情列出实际参测配置。综合榜展示评测覆盖项数和 ECI 的 90% 可信区间。
考查模型准确回答短篇事实问题的能力。
SimpleQA Verified ↗知识问答榜
97 个模型 · 第 1 / 2 页
| 01 | Claude Fable 5.2Anthropic · 2026-11-09 | 94.0% |
|---|---|---|
| 02 | GPT-6.2 AstraOpenAI · 2026-11-17 | 78.5% |
| 03 | Gemini 4 CarbonGoogle DeepMind · 2026-11-10 | 77.7% |
| 04 | Gemini 4 ArgonGoogle DeepMind · 2026-10-13 | 76.0% |
| 05 | GPT-6 AstraOpenAI · 2026-09-03 | 75.6% |
| 06 | Gemini 4 NeonGoogle DeepMind · 2026-11-18 | 75.2% |
| 07 | GPT-6.1 SolOpenAI · 2026-09-29 | 73.9% |
| 08 | Gemini 3.1 ProGoogle DeepMind · 2026-02-19 | 73.5% |
| 09 | Claude Opus 5.5Anthropic · 2026-09-22 | 72.2% |
| 10 | Gemini 4 HydrogenGoogle DeepMind · 2026-11-12 | 70.9% |
| 11 | Claude Fable 5.1Anthropic · 2026-09-01 | 70.8% |
| 12 | Claude Fable 5Anthropic · 2026-06-09 | 70.7% |
| 13 | Gemini 3.8 FlashGoogle DeepMind · 2026-09-02 | 69.7% |
| 13 | GPT-5.6 SolOpenAI · 2026-07-09 | 69.7% |
| 15 | Gemini 3.7 FlashGoogle DeepMind · 2026-08-13 | 69.2% |
| 16 | Gemini 3 FlashGoogle DeepMind · 2025-12-17 | 66.8% |
| 17 | Gemini 3.5 FlashGoogle DeepMind · 2026-05-19 | 66.2% |
| 17 | Gemini 3.6 FlashGoogle DeepMind · 2026-07-21 | 66.2% |
| 19 | GPT-5.5OpenAI · 2026-04-23 | 63.0% |
| 20 | Kimi K3.1 ProMoonshot · 2026-10-23 · 开放权重 | 62.2% |
| 21 | GPT-6 SolOpenAI · 2026-09-22 | 60.7% |
| 22 | Muse Spark 1.2Meta AI · 2026-08-05 | 60.3% |
| 23 | Claude Opus 5Anthropic · 2026-07-24 | 59.9% |
| 24 | Muse Spark 1.1Meta AI · 2026-07-09 | 57.79% |
| 25 | GPT-6.1 TerraOpenAI · 2026-10-26 | 57.3% |
| 26 | Muse Spark OpenMeta AI · 2026-10-15 · 开放权重 | 56.6% |
| 27 | Grok 4.7xAI · 2026-09-21 | 56.0% |
| 28 | Qwen3.7-MaxAlibaba · 2026-05-19 | 55.77% |
| 29 | Claude Opus 4.8Anthropic · 2026-05-28 | 53.0% |
| 30 | DeepSeek V4 Pro 0813DeepSeek · 2026-08-13 · 开放权重 | 52.91% |
| 31 | Qwen 3.6 Max (Preview)Alibaba · 2026-04-20 | 52.0% |
| 32 | Claude Opus 4.7Anthropic · 2026-04-16 | 51.7% |
| 33 | Kimi K3Moonshot · 2026-07-16 · 开放权重 | 50.6% |
| 34 | GPT-5OpenAI · 2025-08-07 | 50.1% |
| 35 | o3OpenAI · 2025-04-16 | 49.4% |
| 36 | Grok 4.6xAI · 2026-08-12 | 49.3% |
| 37 | Qwen3-MaxAlibaba · 2025-09-24 | 48.75% |
| 38 | Grok 4.5xAI · 2026-07-08 | 48.3% |
| 39 | GPT-5.1OpenAI · 2025-11-13 | 48.0% |
| 40 | Qwen3.8 Max (0902)Alibaba · 2026-09-01 | 47.3% |
| 41 | Claude Opus 4.6Anthropic · 2026-02-05 | 47.0% |
| 42 | DeepSeek-V4-ProDeepSeek · 2026-04-24 · 开放权重 | 46.99% |
| 43 | Claude Sonnet 5.5Anthropic · 2026-09-28 | 46.5% |
| 44 | GPT-5.4 ProOpenAI · 2026-03-05 | 46.3% |
| 45 | Qwen 3.8 MaxAlibaba · 2026-08-02 | 45.8% |
| 46 | Claude Opus 4.5Anthropic · 2025-11-24 | 45.7% |
| 47 | GPT-5.4OpenAI · 2026-03-05 | 45.1% |
| 48 | Doubao Seed 2.1 ProByteDance · 2026-06-23 | 44.7% |
| 49 | Qwen 3.6 PlusAlibaba · 2026-03-31 | 44.14% |
| 50 | GPT-5.6 TerraOpenAI · 2026-07-09 | 43.2% |