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.

Overall OCR Diagrams Homework Captioning Entity recognition Creative writing Humor English Chinese
#1 ChineseData Oct 19, 2026
Gemini 3.8 Flash (High)Google · 1416

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 · Chinese

36 models
All Open weights

Sorted 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 · Chinese License
Score Rank Votes
1 GLM-5.3-FlashZ.aiOpen 1363±23 15 699 MIT
2 Kimi K2.6Moonshot AIOpen 1346±20 25 1,027 Modified MIT
3 Kimi K2.5 (Instant)Moonshot AIOpen 1328±34 36 273 Modified MIT
4 DeepSeek-V4.1-Flash-20260910 (Max)DeepSeekOpen 1327±42 37 199 MIT
5 Gemma 4 31BGoogleOpen 1326±14 38 2,734 Apache 2.0
6 MiMo-V2.5XiaomiOpen 1326±17 39 1,580 MIT
7 Qwen3.8-27BAlibabaOpen 1326±30 40 409 Apache 2.0
8 Kimi K2.5 (Thinking)Moonshot AIOpen 1324±16 41 1,767 Modified MIT
9 Qwen3.5-397B-A17BAlibabaOpen 1314±15 43 2,045 Apache 2.0
10 Gemma 4 26B A4BGoogleOpen 1312±17 45 1,609 Apache 2.0
11 MiniMax-M3MiniMaxOpen 1311±20 46 1,048 MiniMax Community License
12 Qwen3.5-122B-A10BAlibabaOpen 1301±20 51 1,011 Apache 2.0
13 Qwen3.5-27BAlibabaOpen 1301±15 52 1,997 Apache 2.0
14 TInkling-SmallThinking MachinesOpen 1292±28 56 461 Apache 2.0
15 Qwen3-VL-235B-A22B-InstructAlibabaOpen 1283±19 62 998 Apache 2.0
16 Mistral Medium 3.5Mistral AIOpen 1244±38 76 287 Modified MIT
17 Mistral Large 3Mistral AIOpen 1240±26 77 569 Apache 2.0
18 Qwen3-VL-235B-A22B-ThinkingAlibabaOpen 1219±43 85 164 Apache 2.0
19 GLM-4.6VZ.aiOpen 1217±41 87 166 MIT
20 Gemma 3 27B ITGoogleOpen 1164±21 96 1,020 Gemma
21 Mistral Small 3.2 (2506)Mistral AIOpen 1163±23 97 847 Apache 2.0
22 Mistral Small 3.1 24B Instruct (2503)Mistral AIOpen 1155±21 99 1,251 Apache 2.0
23 Llama 4 Maverick 17B 128E InstructMetaOpen 1141±34 101 294 Llama 4
24 Qwen2.5-VL-72B-InstructAlibabaOpen 1107±55 105 119 Qwen
25 Llama 4 Scout 17B 16E InstructMetaOpen 1106±36 106 278 Llama
26 Pixtral Large (2411)Mistral AIOpen 1083±51 108 156 MRL
27 Qwen2-VL-72BAlibabaOpen 1072±45 109 291 Qwen
28 OInternVL2 26BOpenGVLabOpen 1065±48 110 287 MIT
29 Qwen2-VL-7B-InstructAlibabaOpen 999±46 116 298 Apache 2.0
30 Llama 3.2 90B Vision InstructMetaOpen 952±43 120 393 Llama 3.2
31 Pixtral 12B (2409)Mistral AIOpen 951±45 121 335 Apache 2.0
32 OInternVL2 4BOpenGVLabOpen 950±55 122 177 MIT
33 Molmo 72B (0924)Ai2Open 936±55 124 182 Apache 2.0
34 LLLaVA 1.6 34BOpen 915±47 126 415 Apache 2.0
35 Llama 3.2 11B Vision InstructMetaOpen 915±49 127 266 Llama 3.2
36 Molmo 7B-D (0924)Ai2Open 890±54 128 171 Apache 2.0

Scores: LMArena leaderboard-dataset (lmarena-ai on Hugging Face, CC-BY-4.0, vision config), Chinese published Oct 19, 2026. Please credit the source when citing.

Sources & methodology

  1. Scores are Bradley-Terry Arena ratings fitted from human votes between two anonymous answers to the same image prompt.
  2. 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.
  3. Fields without public data (price, context length, release date…) are not shown; blanks are –. Nothing is estimated.
  4. 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.