Chatbot Arena (Search): leaderboard
Crowdsourced Elo-rated leaderboard where users compare AI models with web search on real-time information retrieval and grounded citation tasks.
Metric: Arena Score. Source: arena.ai. Saturation forecast: Around September 2027. 34 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.6 Sol (xHigh) | 1257 | #19 (GPT-5.6 Sol) |
| 2 | Claude Opus 4.6 | 1253 | #63 |
| 3 | GPT-5.5 | 1242 | #31 |
| 4 | Claude Opus 4.7 | 1233 | #49 |
| 5 | Claude Fable 5 (High) | 1230 | #16 (Claude Fable 5) |
| 6 | Ernie 5.1 | 1227 | #202 |
| 7 | Claude Sonnet 4.6 | 1221 | #90 |
| 8 | Grok 4.5 | 1213 | #48 |
| 9 | Gemini 3.1 Pro (Preview) | 1210 | #58 |
| 10 | GPT-5.2 | 1207 | #110 |
| 11 | Gemini 3 Pro | 1207 | #80 |
| 12 | Claude Opus 4.8 | 1204 | #37 |
| 13 | Grok 4.20 Multi-Agent | 1204 | #113 |
| 14 | GPT-5.1 | 1199 | #135 |
| 15 | Gemini 3 Flash | 1198 | #99 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=chatbot-arena-search · How It Works · Data refreshed daily, snapshot 2026-10-20.