Chatbot Arena (Vision - Captioning): leaderboard
arena.ai’s vision leaderboard restricted to battles that ask for a caption. Same crowd-vote Arena Score as the headline vision board, refitted over that category’s votes alone.
Metric: Arena Score. Source: arena.ai. 34 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 1282 | #58 |
| 2 | Gemini 3 Pro | 1276 | #80 |
| 3 | GPT-5.2 (High) | 1263 | #110 (GPT-5.2) |
| 4 | Gemini 2.5 Pro | 1251 | #152 |
| 5 | GPT-5.1 (High) | 1248 | #135 (GPT-5.1) |
| 6 | Gemma 4 26B A4B | 1241 | #358 |
| 7 | Gemini 3 Flash | 1234 | #99 |
| 8 | Gemini 3.1 Flash Lite (Preview) | 1227 | #194 |
| 9 | GPT-4o ChatGPT | 1213 | |
| 10 | Gemini 2.5 Flash | 1212 | #245 |
| 11 | Qwen 3 VL 235B A22B Instruct | 1211 | #273 |
| 12 | GPT-4.1 | 1210 | #248 |
| 13 | Kimi K2.5 (Thinking) | 1199 | #144 (Kimi K2.5) |
| 14 | GPT-5 (High) | 1195 | #98 (GPT-5) |
| 15 | GPT-5 Chat | 1194 | #241 |
Interactive version: theaggregate.ai/benchmark?slug=chatbot-arena-vision-captioning · How It Works · Data refreshed daily, snapshot 2026-10-20.