模型排行榜
每日同步
来源日期:2026-11-19
内容更新:2026-11-20 18:05(UTC+8)
联网搜索
联网搜索综合榜,使用官方公开原始评分;竞技场评分越高越好。
| 排名 | 模型 | 机构 | 竞技场评分 | 置信区间 | 投票数 |
|---|---|---|---|---|---|
| 1 | Amazon-Web-Search-Agent | amazon | 1,266.3 | 1,251.4 – 1,281.2 | 1,565 |
| 2 | Qwen3.8-Max | 阿里巴巴 | 1,260.9 | 1,250.2 – 1,271.6 | 3,737 |
| 3 | GPT-5.6 Sol (xHigh) | OpenAI | 1,256.2 | 1,249.7 – 1,262.6 | 68,265 |
| 4 | Claude Opus 4.6 (Search) | Anthropic | 1,251.6 | 1,247.0 – 1,256.3 | 165,940 |
| 5 | GPT-5.5 (Search) | OpenAI | 1,246.4 | 1,241.3 – 1,251.5 | 127,296 |
| 6 | Claude Opus 5 (High) | Anthropic | 1,238.3 | 1,231.6 – 1,245.0 | 39,953 |
| 7 | Claude Fable 5.1 (High) | Anthropic | 1,236.8 | 1,229.8 – 1,243.7 | 39,817 |
| 8 | GPT-6 Astra (Max) | OpenAI | 1,230.5 | 1,222.6 – 1,238.4 | 17,308 |
| 9 | Claude Opus 4.7 | Anthropic | 1,230.0 | 1,225.0 – 1,235.0 | 121,123 |
| 10 | GPT-6.1 Sol (Max) | OpenAI | 1,229.6 | 1,222.3 – 1,237.0 | 33,142 |
| 11 | Claude Fable 5 (High) | Anthropic | 1,227.7 | 1,221.1 – 1,234.4 | 69,017 |
| 12 | ERNIE 5.1 | baidu | 1,227.1 | 1,217.5 – 1,236.7 | 3,912 |
| 13 | Claude Sonnet 4.6 (Search) | Anthropic | 1,220.3 | 1,215.6 – 1,224.9 | 166,671 |
| 14 | Grok 4.5 | xai | 1,215.6 | 1,210.0 – 1,221.1 | 115,851 |
| 15 | Claude Opus 5.5 (Medium) | Anthropic | 1,213.6 | 1,206.7 – 1,220.5 | 37,774 |
| 16 | Gemini 3.1 Pro (Grounding) | 谷歌 | 1,211.4 | 1,206.9 – 1,215.9 | 185,144 |
| 17 | Gemini 3 Pro (Grounding) | 谷歌 | 1,206.8 | 1,201.2 – 1,212.3 | 38,262 |
| 18 | GPT-5.2 (Search) | OpenAI | 1,206.7 | 1,201.1 – 1,212.3 | 54,537 |
| 19 | Grok 4.20 Multi-Agent Beta (0309) | xai | 1,205.2 | 1,199.6 – 1,210.7 | 120,496 |
| 20 | Claude Opus 4.8 | Anthropic | 1,202.3 | 1,196.6 – 1,208.1 | 100,556 |
| 21 | GPT-5.1 (Search) | OpenAI | 1,199.4 | 1,194.1 – 1,204.7 | 61,858 |
| 22 | GPT-5.4 (Search) | OpenAI | 1,198.0 | 1,192.4 – 1,203.6 | 120,996 |
| 23 | Gemini 3 Flash (Grounding) | 谷歌 | 1,195.2 | 1,190.7 – 1,199.6 | 188,717 |
| 24 | GPT-6.2 Astra (Max) | OpenAI | 1,191.4 | 1,171.0 – 1,211.8 | 1,077 |
| 25 | Grok 4.20 Beta 1 | xai | 1,190.8 | 1,184.7 – 1,196.8 | 55,735 |
| 26 | Claude Sonnet 5 (Search) | Anthropic | 1,187.3 | 1,180.9 – 1,193.8 | 70,734 |
| 27 | Claude Opus 4.5 (Search) | Anthropic | 1,179.2 | 1,173.5 – 1,184.8 | 63,735 |
| 28 | GPT-6.2 Sol (Max) | OpenAI | 1,179.2 | 1,159.7 – 1,198.6 | 1,202 |
| 29 | GPT-6 Sol (Max) | OpenAI | 1,173.8 | 1,166.3 – 1,181.3 | 13,355 |
| 30 | GPT-5.2 (Search, Non-Reasoning) | OpenAI | 1,171.9 | 1,166.3 – 1,177.5 | 78,271 |
| 31 | Grok 4.1 Fast (Search) | xai | 1,171.7 | 1,166.7 – 1,176.8 | 84,258 |
| 32 | Grok 4 Fast (Search) | xai | 1,170.8 | 1,166.0 – 1,175.5 | 42,555 |
| 33 | Grok 4.3 | xai | 1,166.1 | 1,161.3 – 1,171.0 | 121,970 |
| 34 | Claude Sonnet 4.5 (Search) | Anthropic | 1,158.8 | 1,153.6 – 1,164.0 | 138,925 |
| 35 | GPT-6 Luna (Max) | OpenAI | 1,156.8 | 1,150.6 – 1,163.0 | 65,680 |
| 36 | Claude Opus 4.1 (Search) | Anthropic | 1,148.2 | 1,143.2 – 1,153.1 | 78,614 |
| 37 | o3 (Search) | OpenAI | 1,144.4 | 1,138.7 – 1,150.1 | 20,616 |
| 38 | Gemini 2.5 Pro (Grounding) | 谷歌 | 1,143.3 | 1,138.5 – 1,148.1 | 85,319 |
| 39 | Grok 4 (Search) | xai | 1,141.7 | 1,135.5 – 1,147.9 | 19,048 |
| 40 | Sonar Reasoning Pro (High) | perplexity | 1,138.8 | 1,132.8 – 1,144.7 | 29,154 |
| 41 | GPT-5 (Search) | OpenAI | 1,133.1 | 1,127.0 – 1,139.2 | 20,707 |
| 42 | Sonar Pro (High) | perplexity | 1,129.6 | 1,123.6 – 1,135.5 | 28,611 |
| 43 | Claude Opus 4 (Search) | Anthropic | 1,126.9 | 1,121.3 – 1,132.4 | 31,209 |
| 44 | Diffbot-Small-XL | diffbot | 1,022.0 | 1,013.4 – 1,030.6 | 6,336 |
| 45 | GPT-4o (Search) | OpenAI | 1,006.2 | 995.0 – 1,017.5 | 3,392 |
数据来源:Arena 官方联网搜索。 使用官方公开数据集,按 CC BY 4.0 标注来源。 各榜单独同步;来源未发布新数据时,内容日期保持不变。
竞技场评分与任务、评测方式有关,不同子榜的分数不宜直接比较。 最近检查:2026-11-22 18:05(UTC+8)。