Vision Model Leaderboard: Best Vision-Language Models
Ranked by LMArena Vision Arena scores: users upload an image, ask a question, and vote between two anonymous models’ answers.
- Top open-weightGemma 4 26B A4B1297Google
In LMArena Vision Arena data published Oct 19, 2026, Anthropic’s Claude Fable 5 (High) ranks #1 of 167 models with 1325 (14,714 votes).
Full leaderboard · Captioning
34 modelsSorted by LMArena score by default; ± is half the 95% confidence interval. Category columns show the model’s score on that sub-board; – means no score.
| # | Model | LMArena · Captioning | License | ||
|---|---|---|---|---|---|
| Score↑ | Rank | Votes | |||
| 1 |  Mistral Small 3.2 (2506)Mistral AIOpen |
1059±49 | 34 | 177 | Apache 2.0 |
| 2 | Mistral Medium 3 (2505)Mistral AI |
1094±49 | 33 | 160 | Proprietary |
| 3 | Gemini 2.0 Flash (001)Google |
1130±61 | 32 | 106 | Proprietary |
| 4 | Mistral Medium 3.1 (2508)Mistral AI |
1145±30 | 31 | 403 | Proprietary |
| 5 | Gemma 3 27B ITGoogleOpen |
1151±39 | 30 | 258 | Gemma |
| 6 | Mistral Small 3.1 24B Instruct (2503)Mistral AIOpen |
1153±38 | 29 | 258 | Apache 2.0 |
| 7 |  GPT-5.1OpenAI |
1165±49 | 28 | 123 | Proprietary |
| 8 |  Qwen3.5-27BAlibabaOpen |
1170±81 | 27 | 61 | Apache 2.0 |
| 9 | GPT-4.1 mini (2025-04-14)OpenAI |
1174±32 | 25 | 394 | Proprietary |
| 10 | GPT-5.2OpenAI |
1174±55 | 26 | 107 | Proprietary |
| 11 | Gemini 2.5 Flash-Lite Preview (Thinking, 06-17)Google |
1184±32 | 24 | 384 | Proprietary |
| 12 | o4-mini (2025-04-16)OpenAI |
1204±30 | 22 | 423 | Proprietary |
| 13 | GPT-5 (High)OpenAI |
1204±32 | 23 | 356 | Proprietary |
| 14 | GPT-4.1 (2025-04-14)OpenAI |
1207±31 | 21 | 422 | Proprietary |
| 15 | o3 (2025-04-16)OpenAI |
1210±28 | 20 | 540 | Proprietary |
| 16 | GPT-5 ChatOpenAI |
1212±32 | 19 | 375 | Proprietary |
| 17 | GPT-5 mini (High)OpenAI |
1217±37 | 18 | 283 | Proprietary |
| 18 | Gemma 4 31BGoogleOpen |
1218±54 | 17 | 106 | Apache 2.0 |
| 19 | ChatGPT-4o Latest (20250326)OpenAI |
1224±36 | 16 | 273 | Proprietary |
| 20 | Qwen3.5-397B-A17BAlibabaOpen |
1230±57 | 15 | 78 | Apache 2.0 |
| 21 | Gemini 3 Flash (Minimal)Google |
1241±43 | 14 | 173 | Proprietary |
| 22 |  Grok 4 (0709)xAI |
1245±34 | 13 | 351 | Proprietary |
| 23 | GPT-5.1 (High)OpenAI |
1252±55 | 12 | 101 | Proprietary |
| 24 |  Kimi K2.5 (Thinking)Moonshot AIOpen |
1257±48 | 11 | 125 | Modified MIT |
| 25 | Gemini 3.1 Flash-Lite PreviewGoogle |
1259±64 | 10 | 73 | Proprietary |
| 26 |  GLM-5V-TurboZ.ai |
1264±73 | 9 | 53 | Proprietary |
| 27 | GPT-5.2 (High)OpenAI |
1267±58 | 8 | 94 | Proprietary |
| 28 | Gemini 2.5 FlashGoogle |
1277±24 | 7 | 582 | Proprietary |
| 29 | Qwen3-VL-235B-A22B-InstructAlibabaOpen |
1281±45 | 5 | 146 | Apache 2.0 |
| 30 | Gemini 3 FlashGoogle |
1281±39 | 6 | 191 | Proprietary |
| 31 | Gemma 4 26B A4BGoogleOpen |
1297±70 | 4 | 63 | Apache 2.0 |
| 32 | Gemini 2.5 ProGoogle |
1304±23 | 3 | 796 | Proprietary |
| 33 | Gemini 3 ProGoogle |
1325±35 | 2 | 218 | Proprietary |
| 34 | Gemini 3.1 Pro PreviewGoogle |
1345±49 | 1 | 137 | Proprietary |
Scores: LMArena leaderboard-dataset (lmarena-ai on Hugging Face, CC-BY-4.0, vision config), Captioning published Oct 19, 2026. Please credit the source when citing.
Sources & methodology
- LMLMArena leaderboard dataset (Hugging Face, CC-BY-4.0)
Vision config, published Oct 19, 2026; 167 models on the overall board.
- Scores are Bradley-Terry Arena ratings fitted from human votes between two anonymous answers to the same image prompt.
- Categories (OCR, diagrams, homework, captioning, Chinese/English…) are separate LMArena boards computed on prompts of that type, not a breakdown of the overall score; ranks can differ a lot.
- Fields without public data (price, context length, release date…) are not shown; blanks are –. Nothing is estimated.
- Synced daily. JSON API: /api/public/vision-leaderboard.php
This week
Collecting data: 1/7 days recorded; weekly changes appear after 7 days
Synced automatically once a day; ranks are compared only between real snapshots, never backfilled.FAQ
What is the best vision-language model right now?
In LMArena Vision Arena data published Oct 19, 2026, Anthropic’s Claude Fable 5 (High) ranks #1 of 167 models with 1325 (14,714 votes).
What is each vendor’s best vision model?
Top overall entry per vendor: Anthropic Claude Fable 5 (High) (1325, #1), Google Gemini 3.8 Flash (High) (1315, #7), Alibaba Qwen3.8-Max (1313, #10), Meta Muse Spark 1.3 (Max) (1308, #12), OpenAI GPT-5.4 (High) (1303, #18).
Which model is best at OCR?
On LMArena’s OCR board (123 models), Claude Fable 5 (High) is #1 with 1337.
What is the best open-weight vision model?
50 open-weight models are listed; the top one is GLM-5.3-Flash (1294, MIT).
How is this different from benchmarks like MMMU?
MMMU and similar benchmarks auto-score a fixed question set; this board comes from blind human votes on people’s own images and questions, reflecting real-use preference. They are not convertible.