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 HomeworkData Oct 19, 2026
Claude Opus 5.5 (High)Anthropic · 1362

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 · Homework

118 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 · Homework License
Score Rank Votes
1 Molmo 2 8BAi2Open 1069±50 118 187 Apache 2.0
2 Claude 3.5 Haiku (20241022)Anthropic 1119±39 117 212 Proprietary
3 Claude 3.5 Sonnet (20241022)Anthropic 1136±39 116 210 Proprietary
4 Llama 4 Scout 17B 16E InstructMetaOpen 1148±26 115 477 Llama
5 GLM-4.6VZ.aiOpen 1155±36 113 271 MIT
6 SStep-1o Turbo (202506)StepFun 1155±36 114 237 Proprietary
7 GPT-5 nano (High)OpenAI 1157±35 112 260 Proprietary
8 Mistral Small 3.1 24B Instruct (2503)Mistral AIOpen 1158±21 111 809 Apache 2.0
9 Gemma 3 27B ITGoogleOpen 1160±19 110 1,034 Gemma
10 GLM-4.5VZ.aiOpen 1163±38 109 215 MIT
11 Llama 4 Maverick 17B 128E InstructMetaOpen 1164±25 108 526 Llama 4
12 Mistral Medium 3.1 (2508)Mistral AI 1165±13 106 2,146 Proprietary
13 Mistral Medium 3 (2505)Mistral AI 1165±21 107 809 Proprietary
14 Grok 4 (0709)xAI 1170±19 105 1,075 Proprietary
15 Grok 4.1 Fast (Reasoning)xAI 1172±14 104 2,064 Proprietary
16 Mistral Small 3.2 (2506)Mistral AIOpen 1173±24 103 614 Apache 2.0
17 Gemini 2.0 Flash (001)Google 1175±22 102 763 Proprietary
18 SStep-3StepFunOpen 1183±39 101 215 Apache 2.0
19 Gemini 2.5 Flash-Lite Preview (Thinking, 06-17)Google 1190±17 100 1,327 Proprietary
20 Claude 3.7 Sonnet (Thinking 32K, 20250219)Anthropic 1192±42 99 171 Proprietary
21 Claude Opus 4 (Thinking 16K, 20250514)Anthropic 1197±45 98 161 Proprietary
22 Claude 3.7 Sonnet (20250219)Anthropic 1206±37 97 217 Proprietary
23 GPT-5 mini (High)OpenAI 1210±19 96 976 Proprietary
24 GPT-4.1 mini (2025-04-14)OpenAI 1214±17 95 1,310 Proprietary
25 Claude Sonnet 4 (20250514)Anthropic 1215±36 93 230 Proprietary
26 Gemini 2.5 Flash-Lite Preview (Non-Thinking, 09-2025)Google 1215±22 94 605 Proprietary
27 Mistral Large 3Mistral AIOpen 1216±20 92 977 Apache 2.0
28 GPT-5.4 nano (High)OpenAI 1225±11 91 3,429 Proprietary
29 GPT-5 (High)OpenAI 1227±16 90 1,409 Proprietary
30 Claude Opus 4 (20250514)Anthropic 1228±33 89 306 Proprietary
31 NNVIDIA-Nemotron-3.5-Super-VL-120B-A12B (NVFP4)NVIDIAOpen 1229±32 88 380 OpenMDW-1.1
32 Mistral Medium 3.5Mistral AIOpen 1230±22 87 803 Modified MIT
33 Qwen3-VL-235B-A22B-ThinkingAlibabaOpen 1234±32 84 298 Apache 2.0
34 o4-mini (2025-04-16)OpenAI 1234±16 85 1,588 Proprietary
35 o3 (2025-04-16)OpenAI 1234±15 86 1,930 Proprietary
36 Grok 4.3xAI 1241±11 82 3,560 Proprietary
37 Hunyuan Vision 1.5 ThinkingTencent 1241±33 83 297 Proprietary
38 GPT-5.1OpenAI 1243±15 81 1,417 Proprietary
39 Gemini 2.5 FlashGoogle 1244±9 80 4,863 Proprietary
40 GPT-5.5 InstantOpenAI 1247±18 79 1,173 Proprietary
41 GPT-4.1 (2025-04-14)OpenAI 1248±16 78 1,478 Proprietary
42 ChatGPT-4o Latest (20250326)OpenAI 1252±13 77 2,120 Proprietary
43 ERNIE-5.0-Preview (1220)Baidu 1253±29 75 363 Proprietary
44 MiMo-V2-OmniXiaomi 1253±15 76 1,629 Proprietary
45 Qwen3.5-27BAlibabaOpen 1256±10 74 3,843 Apache 2.0
46 TInkling-SmallThinking MachinesOpen 1257±20 73 835 Apache 2.0
47 GPT-5.2OpenAI 1258±11 72 3,018 Proprietary
48 Qwen3.5-122B-A10BAlibabaOpen 1261±14 70 1,929 Apache 2.0
49 GPT-6 Luna (Max)OpenAI 1261±25 71 550 Proprietary
50 MiMo-V2.5XiaomiOpen 1262±10 69 4,076 MIT
51 MiMo-V2.6-ProXiaomiOpen 1264±39 67 226 MIT
52 Kimi K2.5 (Instant)Moonshot AIOpen 1264±23 68 573 Modified MIT
53 SStep 5 PreviewStepFun 1265±38 66 233 Proprietary
54 MiniMax-M3MiniMaxOpen 1268±13 65 2,405 MiniMax Community License
55 Grok 4.20 Beta (Reasoning, 0309)xAI 1269±11 64 3,619 Proprietary
56 Gemini 3.1 Flash-Lite PreviewGoogle 1271±10 63 3,947 Proprietary
57 GPT-5.1 (High)OpenAI 1272±17 62 1,218 Proprietary
58 GLM-5V-TurboZ.ai 1273±9 61 5,226 Proprietary
59 Grok 4.20 Multi-Agent Beta (0309)xAI 1275±11 60 3,348 Proprietary
60 GPT-5.2 (High)OpenAI 1276±12 59 2,888 Proprietary
61 Qwen3-VL-235B-A22B-InstructAlibabaOpen 1277±14 58 1,611 Apache 2.0
62 GPT-5 ChatOpenAI 1278±17 57 1,325 Proprietary
63 Gemini 2.5 ProGoogle 1281±9 56 5,415 Proprietary
64 GPT-5.6 Luna (xHigh)OpenAI 1282±18 55 1,128 Proprietary
65 Grok 4.6 (High)xAI 1283±20 52 873 Proprietary
66 Gemini 2.5 Flash Preview (09-2025)Google 1283±23 53 604 Proprietary
67 MiMo-V2.6-FlashXiaomiOpen 1283±26 54 546 MIT
68 Dola Seed 2.0 ProByteDance 1284±14 51 1,842 Proprietary
69 GPT-5.4 mini (High)OpenAI 1285±11 50 3,492 Proprietary
70 GPT-6 Sol (Max)OpenAI 1286±26 48 518 Proprietary
71 GPT-5.2 Chat (20260210)OpenAI 1286±12 49 2,485 Proprietary
72 Muse Spark 1.1Meta 1287±16 47 1,354 Proprietary
73 Gemini 3.5 Flash-LiteGoogle 1291±19 46 1,010 Proprietary
74 Qwen3.5-397B-A17BAlibabaOpen 1292±10 45 4,001 Apache 2.0
75 Gemma 4 26B A4BGoogleOpen 1293±10 44 3,822 Apache 2.0
76 Grok 4.5xAI 1295±17 41 1,284 Proprietary
77 Qwen3.7-PlusAlibaba 1295±15 42 1,646 Proprietary
78 Qwen3.8-27BAlibabaOpen 1295±20 43 919 Apache 2.0
79 Muse Spark 1.3 (xHigh)Meta 1296±23 39 680 Proprietary
80 Muse Spark 1.3 (Max)Meta 1296±21 40 752 Proprietary
81 Kimi K2.5 (Thinking)Moonshot AIOpen 1297±10 38 3,649 Modified MIT
82 Gemini 3 Flash (Minimal)Google 1298±9 37 5,051 Proprietary
83 Muse SparkMeta 1299±19 36 945 Proprietary
84 GPT-5.6 Sol (xHigh)OpenAI 1305±17 35 1,190 Proprietary
85 DeepSeek-V4.1-Flash-20260910 (Max)DeepSeekOpen 1307±25 34 521 MIT
86 GPT-5.4OpenAI 1309±12 32 3,080 Proprietary
87 Grok 4.7 (xHigh)xAI 1309±25 33 517 Proprietary
88 Kimi K2.6Moonshot AIOpen 1310±13 30 2,263 Modified MIT
89 Gemma 4 31BGoogleOpen 1310±9 31 6,435 Apache 2.0
90 Claude Sonnet 4.6Anthropic 1311±11 28 3,700 Proprietary
91 Claude Sonnet 5 (High)Anthropic 1311±16 29 1,500 Proprietary
92 GPT-5.6 Terra (xHigh)OpenAI 1312±17 27 1,286 Proprietary
93 Gemini 3 FlashGoogle 1313±9 26 5,232 Proprietary
94 Gemini 3.6 Flash (High)Google 1314±19 25 993 Proprietary
95 Gemini 3 ProGoogle 1315±14 24 1,779 Proprietary
96 Gemini 3.1 Pro PreviewGoogle 1316±8 23 6,293 Proprietary
97 GLM-5.3-FlashZ.aiOpen 1318±13 22 2,229 MIT
98 Gemini 3.7 Flash (High)Google 1320±23 21 665 Proprietary
99 Gemini 3.5 Flash (High)Google 1321±15 20 1,641 Proprietary
100 Claude Opus 4.6Anthropic 1327±11 18 3,688 Proprietary
101 GPT-5.5OpenAI 1327±11 19 3,602 Proprietary
102 Claude Opus 4.8Anthropic 1328±13 17 2,232 Proprietary
103 Gemini 3.5 Flash (Medium)Google 1330±15 16 1,553 Proprietary
104 Claude Sonnet 5.5 (xHigh)Anthropic 1331±30 15 395 Proprietary
105 GPT-5.5 (High)OpenAI 1332±11 14 3,535 Proprietary
106 Claude Opus 4.7Anthropic 1335±12 13 3,157 Proprietary
107 GPT-5.4 (High)OpenAI 1336±10 11 4,003 Proprietary
108 Qwen3.8-MaxAlibaba 1336±19 12 909 Proprietary
109 GPT-6 Astra (Max)OpenAI 1337±25 7 560 Proprietary
110 Claude Opus 4.7 (High)Anthropic 1337±12 8 3,084 Proprietary
111 Claude Opus 4.8 (High)Anthropic 1337±13 9 2,077 Proprietary
112 Claude Opus 4.6 (High)Anthropic 1337±12 10 2,845 Proprietary
113 Claude Opus 5 (High)Anthropic 1338±14 6 1,899 Proprietary
114 Claude Fable 5.1 (Max)Anthropic 1345±21 5 756 Proprietary
115 Gemini 3.8 Flash (High)Google 1351±22 4 681 Proprietary
116 GPT-6.1 Sol (Max)OpenAI 1356±31 3 368 Proprietary
117 Claude Fable 5 (High)Anthropic 1361±15 2 1,751 Proprietary
118 Claude Opus 5.5 (High)Anthropic 1362±29 1 422 Proprietary

Scores: LMArena leaderboard-dataset (lmarena-ai on Hugging Face, CC-BY-4.0, vision config), Homework 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.