Chatbot Arena (Vision - Captioning): leaderboard
Metric: Arena Score. Source: arena.ai. 32 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 1273 |
| 2 | Gemini 3 Pro | 1266 |
| 3 | GPT-5.2 (High) | 1264 |
| 4 | Gemini 2.5 Pro | 1250 |
| 5 | Gemini 3.1 Flash Lite (Preview) | 1239 |
| 6 | GPT-5.1 (High) | 1238 |
| 7 | Gemma 4 26B A4B | 1228 |
| 8 | Gemini 3 Flash | 1225 |
| 9 | Gemini 2.5 Flash | 1213 |
| 10 | GPT-4o ChatGPT | 1212 |
| 11 | GPT-4.1 | 1209 |
| 12 | Qwen 3 VL 235B A22B Instruct | 1209 |
| 13 | Kimi K2.5 (Thinking) | 1198 |
| 14 | GPT-5 Chat | 1198 |
| 15 | GPT-4.1 Mini | 1193 |
Interactive version: theaggregate.ai/benchmark?slug=chatbot-arena-vision-captioning · How It Works · Data refreshed daily, snapshot 2026-09-05.