Arena Vision — Captioning (No Style Control)
Multimodal
Raw Vision Arena rating for image-captioning prompts without style-control adjustment. Higher ratings indicate stronger preference in blind pairwise user votes.
Top models (higher is better)
| Model | Score |
|---|---|
| Gemini 3.1 Pro Preview | 1345 |
| Gemini 3 Pro Preview | 1325 |
| Gemini 2.5 Pro | 1304 |
| Gemma 4 26B A4B IT | 1297 |
| Gemini 3 Flash Preview | 1281 |
| Gemini 2.5 Flash | 1277 |
| GPT-5.2 | 1267 |
| GLM 5V Turbo | 1264 |
| Gemini 3.1 Flash-Lite Preview | 1259 |
| Kimi K2.5 | 1257 |
| GPT-5.1 | 1252 |
| Grok 4 | 1245 |
| Qwen3.5 397B A17B | 1230 |
| Gemma 4 31B IT | 1218 |
| GPT-5 Mini | 1217 |
| o3 | 1210 |
| GPT-4.1 | 1207 |
| O4 Mini | 1204 |
| GPT-5 | 1204 |
| Gemini 2.5 Flash-Lite | 1184 |
| GPT-4.1 Mini | 1174 |
| Mistral Small 3.1 24B Instruct 2503 | 1153 |
| Gemma 3 27B IT | 1151 |
| Mistral Medium 3.1 | 1145 |
| Gemini 2.0 Flash | 1130 |
| Mistral Medium 3 | 1094 |
| Mistral Small 3.2 24B Instruct 2506 | 1059 |