Multimodal · Elo
LMArena Vision
Arena for image-understanding chats: users upload images, compare two anonymous models and vote.
What it measures
Human preference on image + text questions.
Limitations
Fewer votes than the text arena, so wider confidence intervals.
Contamination risk: Low
Live prompts.
Leaderboard
Top 100 of 143
| # | Model | Provider | Score | Rank | Setting | Source type | Date |
|---|---|---|---|---|---|---|---|
| 1 | Claude Fable 5 | Anthropic | 1309 | 1 | Overall, style control on; 13709 votes; 95% CI 1301–1316“claude-fable-5-high” | Official leaderboard | Oct 2, 2026 |
| 2 | Qwen 3.8 Max | Alibaba (Qwen) | 1301 | 2 | Overall, style control on; 10040 votes; 95% CI 1294–1309“qwen3.8-max” | Official leaderboard | Oct 2, 2026 |
| 3 | Claude Opus 4.6 | Anthropic | 1299 | 3 | Overall, style control on; 22328 votes; 95% CI 1293–1306“claude-opus-4-6-high” | Official leaderboard | Oct 2, 2026 |
| 4 | Claude Opus 4.7 | Anthropic | 1298 | 4 | Overall, style control on; 23169 votes; 95% CI 1292–1305“claude-opus-4-7” | Official leaderboard | Oct 2, 2026 |
| 5 | Gemini 3.7 Flash | 1296 | 6 | Overall, style control on; 3814 votes; 95% CI 1286–1306“gemini-3.7-flash-high” | Official leaderboard | Oct 2, 2026 | |
| 6 | Muse Spark | Meta | 1294 | 8 | Overall, style control on; 5868 votes; 95% CI 1284–1303“muse-spark” | Official leaderboard | Oct 2, 2026 |
| 7 | muse-spark-1.2 (xHigh) | Meta | 1293 | 9 | Overall, style control on; 1951 votes; 95% CI 1279–1307 | Official leaderboard | Oct 2, 2026 |
| 8 | gpt-6.1-sol-max | OpenAI | 1291 | 10 | Overall, style control on; 1408 votes; 95% CI 1275–1307 | Official leaderboard | Oct 2, 2026 |
| 9 | Gemini 3.8 Flash | 1290 | 11 | Overall, style control on; 3086 votes; 95% CI 1279–1302“gemini-3.8-flash-high” | Official leaderboard | Oct 2, 2026 | |
| 10 | muse-spark-1.3-max | Meta | 1290 | 12 | Overall, style control on; 3843 votes; 95% CI 1280–1301 | Official leaderboard | Oct 2, 2026 |
| 11 | Gemini 3 Pro | 1289 | 13 | Overall, style control on; 13620 votes; 95% CI 1282–1297“gemini-3-pro” | Official leaderboard | Oct 2, 2026 | |
| 12 | Claude Opus 5 | Anthropic | 1289 | 14 | Overall, style control on; 15900 votes; 95% CI 1282–1296“claude-opus-5-high” | Official leaderboard | Oct 2, 2026 |
| 13 | claude-fable-5.1-max | Anthropic | 1288 | 15 | Overall, style control on; 4312 votes; 95% CI 1278–1298 | Official leaderboard | Oct 2, 2026 |
| 14 | GPT-5.5 | OpenAI | 1287 | 16 | Overall, style control on; 27223 votes; 95% CI 1281–1292“gpt-5.5” | Official leaderboard | Oct 2, 2026 |
| 15 | Claude Opus 4.8 | Anthropic | 1286 | 17 | Overall, style control on; 18350 votes; 95% CI 1280–1293“claude-opus-4-8-high” | Official leaderboard | Oct 2, 2026 |
| 16 | GPT-5.6 Sol | OpenAI | 1285 | 18 | Overall, style control on; 10421 votes; 95% CI 1277–1292“gpt-5.6-sol-xhigh” | Official leaderboard | Oct 2, 2026 |
| 17 | Gemini 3.5 Flash | 1284 | 19 | Overall, style control on; 12467 votes; 95% CI 1277–1291“gemini-3.5-flash-medium” | Official leaderboard | Oct 2, 2026 | |
| 18 | gpt-6-astra-max | OpenAI | 1284 | 20 | Overall, style control on; 3175 votes; 95% CI 1273–1296 | Official leaderboard | Oct 2, 2026 |
| 19 | GPT-5.4 | OpenAI | 1284 | 22 | Overall, style control on; 29177 votes; 95% CI 1278–1290“gpt-5.4-high” | Official leaderboard | Oct 2, 2026 |
| 20 | Muse Spark 1.1 | Meta | 1281 | 23 | Overall, style control on; 11255 votes; 95% CI 1274–1288“muse-spark-1.1” | Official leaderboard | Oct 2, 2026 |
| 21 | Gemini 3.6 Flash | 1280 | 25 | Overall, style control on; 7392 votes; 95% CI 1271–1288“gemini-3.6-flash-high” | Official leaderboard | Oct 2, 2026 | |
| 22 | Gemini 3.1 Pro | 1279 | 27 | Overall, style control on; 45558 votes; 95% CI 1274–1285“gemini-3.1-pro-preview” | Official leaderboard | Oct 2, 2026 | |
| 23 | Grok 4.5 | xAI | 1279 | 28 | Overall, style control on; 11039 votes; 95% CI 1272–1286“grok-4.5” | Official leaderboard | Oct 2, 2026 |
| 24 | GLM-5.3-FlashOW | Zhipu AI (Z.ai) | 1278 | 30 | Overall, style control on; 5479 votes; 95% CI 1269–1288“glm-5.3-flash” | Official leaderboard | Oct 2, 2026 |
| 25 | gpt-5.2-chat-latest | OpenAI | 1278 | 31 | Overall, style control on; 16480 votes; 95% CI 1271–1285“gpt-5.2-chat-latest-20260210” | Official leaderboard | Oct 2, 2026 |
| 26 | GPT-5.5 Instant | OpenAI | 1276 | 32 | Overall, style control on; 7605 votes; 95% CI 1267–1285“gpt-5.5-instant” | Official leaderboard | Oct 2, 2026 |
| 27 | Claude Sonnet 4.6 | Anthropic | 1275 | 33 | Overall, style control on; 27429 votes; 95% CI 1269–1281“claude-sonnet-4-6” | Official leaderboard | Oct 2, 2026 |
| 28 | Gemini 3 Flash | 1273 | 34 | Overall, style control on; 39142 votes; 95% CI 1268–1278“gemini-3-flash” | Official leaderboard | Oct 2, 2026 | |
| 29 | Claude Sonnet 5.5 | Anthropic | 1268 | 35 | Overall, style control on; 1608 votes; 95% CI 1253–1284“claude-sonnet-5.5-xhigh” | Official leaderboard | Oct 2, 2026 |
| 30 | GPT-5.6 Terra | OpenAI | 1268 | 36 | Overall, style control on; 10292 votes; 95% CI 1261–1276“gpt-5.6-terra-xhigh” | Official leaderboard | Oct 2, 2026 |
| 31 | Qwen3.7-Plus | Alibaba (Qwen) | 1266 | 37 | Overall, style control on; 13065 votes; 95% CI 1259–1272“qwen3.7-plus” | Official leaderboard | Oct 2, 2026 |
| 32 | Kimi K2.6OW | Moonshot AI | 1265 | 38 | Overall, style control on; 16387 votes; 95% CI 1258–1272“kimi-k2.6” | Official leaderboard | Oct 2, 2026 |
| 33 | Claude Sonnet 5 | Anthropic | 1265 | 39 | Overall, style control on; 13029 votes; 95% CI 1257–1271“claude-sonnet-5-high” | Official leaderboard | Oct 2, 2026 |
| 34 | Gemini 3.5 Flash-Lite | 1264 | 40 | Overall, style control on; 7579 votes; 95% CI 1256–1273“gemini-3.5-flash-lite” | Official leaderboard | Oct 2, 2026 | |
| 35 | Grok 4.6 | xAI | 1264 | 41 | Overall, style control on; 5736 votes; 95% CI 1254–1273“grok-4.6-high” | Official leaderboard | Oct 2, 2026 |
| 36 | gemma-4-31bOW | 1262 | 42 | Overall, style control on; 39111 votes; 95% CI 1256–1268 | Official leaderboard | Oct 2, 2026 | |
| 37 | GPT-5.6 Luna | OpenAI | 1260 | 44 | Overall, style control on; 10446 votes; 95% CI 1252–1267“gpt-5.6-luna-xhigh” | Official leaderboard | Oct 2, 2026 |
| 38 | Step 5 Preview | StepFun | 1260 | 45 | Overall, style control on; 604 votes; 95% CI 1235–1284 | Official leaderboard | Oct 2, 2026 |
| 39 | dola-seed-2.0-pro | ByteDance | 1257 | 46 | Overall, style control on; 10981 votes; 95% CI 1249–1264 | Official leaderboard | Oct 2, 2026 |
| 40 | grok-4.20-beta-0309-reasoning | xAI | 1255 | 47 | Overall, style control on; 26832 votes; 95% CI 1249–1262 | Official leaderboard | Oct 2, 2026 |
| 41 | Kimi K2.5OW | Moonshot AI | 1252 | 48 | Overall, style control on; 28326 votes; 95% CI 1246–1258“kimi-k2.5-thinking” | Official leaderboard | Oct 2, 2026 |
| 42 | GPT-5.1 | OpenAI | 1251 | 49 | Overall, style control on; 9842 votes; 95% CI 1244–1259“gpt-5.1-high” | Official leaderboard | Oct 2, 2026 |
| 43 | GPT-5.4 Mini | OpenAI | 1251 | 50 | Overall, style control on; 25734 votes; 95% CI 1244–1257“gpt-5.4-mini-high” | Official leaderboard | Oct 2, 2026 |
| 44 | grok-4.20-multi-agent-beta-0309 | xAI | 1250 | 51 | Overall, style control on; 24571 votes; 95% CI 1244–1256 | Official leaderboard | Oct 2, 2026 |
| 45 | Gemini 2.5 Pro (Jun 2025) | 1248 | 52 | Overall, style control on; 88813 votes; 95% CI 1243–1252“gemini-2.5-pro” | Official leaderboard | Oct 2, 2026 | |
| 46 | MiMo-V2.6 FlashOW | Xiaomi Corp | 1248 | 53 | Overall, style control on; 2275 votes; 95% CI 1234–1261“mimo-v2.6-flash” | Official leaderboard | Oct 2, 2026 |
| 47 | Qwen3.5 397B-A17BOW | Alibaba (Qwen) | 1247 | 54 | Overall, style control on; 31441 votes; 95% CI 1242–1252“qwen3.5-397b-a17b” | Official leaderboard | Oct 2, 2026 |
| 48 | MiMo-V2.6 ProOW | Xiaomi Corp | 1245 | 55 | Overall, style control on; 1787 votes; 95% CI 1231–1260“mimo-v2.6-pro” | Official leaderboard | Oct 2, 2026 |
| 49 | Grok 4.3 Beta | xAI | 1243 | 56 | Overall, style control on; 26300 votes; 95% CI 1237–1249“grok-4.3” | Official leaderboard | Oct 2, 2026 |
| 50 | GPT-5.2 | OpenAI | 1243 | 57 | Overall, style control on; 19039 votes; 95% CI 1237–1249“gpt-5.2-high” | Official leaderboard | Oct 2, 2026 |
| 51 | Qwen 3.8 27BOW | Alibaba (Qwen) | 1243 | 58 | Overall, style control on; 6422 votes; 95% CI 1234–1251“qwen3.8-27b” | Official leaderboard | Oct 2, 2026 |
| 52 | Gemma 4 26B A4BOW | 1242 | 59 | Overall, style control on; 23954 votes; 95% CI 1235–1248“gemma-4-26b-a4b” | Official leaderboard | Oct 2, 2026 | |
| 53 | chatgpt-4o-latest | OpenAI | 1239 | 61 | Overall, style control on; 23416 votes; 95% CI 1233–1244“chatgpt-4o-latest-20250326” | Official leaderboard | Oct 2, 2026 |
| 54 | MiniMax-M3OW | MiniMax | 1237 | 63 | Overall, style control on; 18308 votes; 95% CI 1230–1243“minimax-m3” | Official leaderboard | Oct 2, 2026 |
| 55 | Gemini-3.1-Flash-Lite-Preview | 1235 | 64 | Overall, style control on; 27702 votes; 95% CI 1229–1240“gemini-3.1-flash-lite-preview” | Official leaderboard | Oct 2, 2026 | |
| 56 | MiMo-V2.5OW | Xiaomi Corp | 1233 | 65 | Overall, style control on; 28160 votes; 95% CI 1227–1239“mimo-v2.5” | Official leaderboard | Oct 2, 2026 |
| 57 | Gemini 2.5 Flash (Sep 2025) | 1230 | 66 | Overall, style control on; 4794 votes; 95% CI 1220–1240“gemini-2.5-flash-preview-09-2025” | Official leaderboard | Oct 2, 2026 | |
| 58 | glm-5v-turbo | Zhipu AI (Z.ai) | 1230 | 67 | Overall, style control on; 41533 votes; 95% CI 1225–1236 | Official leaderboard | Oct 2, 2026 |
| 59 | Qwen3.5-122B-A10BOW | Alibaba (Qwen) | 1226 | 69 | Overall, style control on; 13242 votes; 95% CI 1220–1233“qwen3.5-122b-a10b” | Official leaderboard | Oct 2, 2026 |
| 60 | GPT-5 | OpenAI | 1226 | 70 | Overall, style control on; 39910 votes; 95% CI 1219–1233“gpt-5-chat” | Official leaderboard | Oct 2, 2026 |
| 61 | GPT-4.5 | OpenAI | 1225 | 71 | Overall, style control on; 2925 votes; 95% CI 1213–1237“gpt-4.5-preview-2025-02-27” | Official leaderboard | Oct 2, 2026 |
| 62 | ernie-5.0-preview-1220 | Baidu | 1219 | 72 | Overall, style control on; 3352 votes; 95% CI 1208–1230 | Official leaderboard | Oct 2, 2026 |
| 63 | Qwen3.5-27BOW | Alibaba (Qwen) | 1216 | 73 | Overall, style control on; 28993 votes; 95% CI 1211–1222“qwen3.5-27b” | Official leaderboard | Oct 2, 2026 |
| 64 | mimo-v2-omni | Xiaomi | 1216 | 74 | Overall, style control on; 10665 votes; 95% CI 1209–1223 | Official leaderboard | Oct 2, 2026 |
| 65 | o3 | OpenAI | 1216 | 75 | Overall, style control on; 45593 votes; 95% CI 1209–1222“o3-2025-04-16” | Official leaderboard | Oct 2, 2026 |
| 66 | Gemini 2.5 Flash (Jun 2025) | 1215 | 76 | Overall, style control on; 60113 votes; 95% CI 1210–1220“gemini-2.5-flash” | Official leaderboard | Oct 2, 2026 | |
| 67 | qwen3-vl-235b-a22b-instructOW | Alibaba (Qwen) | 1215 | 77 | Overall, style control on; 12728 votes; 95% CI 1208–1222 | Official leaderboard | Oct 2, 2026 |
| 68 | GPT-4.1 | OpenAI | 1214 | 78 | Overall, style control on; 41020 votes; 95% CI 1207–1221“gpt-4.1-2025-04-14” | Official leaderboard | Oct 2, 2026 |
| 69 | Claude Opus 4 | Anthropic | 1207 | 80 | Overall, style control on; 1290 votes; 95% CI 1191–1223“claude-opus-4-20250514-thinking-16k” | Official leaderboard | Oct 2, 2026 |
| 70 | Inkling-SmallOW | Thinking Machines | 1206 | 81 | Overall, style control on; 6904 votes; 95% CI 1198–1215“Inkling Small” | Official leaderboard | Oct 2, 2026 |
| 71 | Claude Sonnet 4 | Anthropic | 1206 | 82 | Overall, style control on; 1145 votes; 95% CI 1189–1223“claude-sonnet-4-20250514-thinking-32k” | Official leaderboard | Oct 2, 2026 |
| 72 | o4-mini | OpenAI | 1201 | 83 | Overall, style control on; 41340 votes; 95% CI 1194–1208“o4-mini-2025-04-16” | Official leaderboard | Oct 2, 2026 |
| 73 | GPT-4.1 mini | OpenAI | 1201 | 84 | Overall, style control on; 40223 votes; 95% CI 1193–1208“gpt-4.1-mini-2025-04-14” | Official leaderboard | Oct 2, 2026 |
| 74 | mistral-large-3OW | Mistral AI | 1201 | 85 | Overall, style control on; 8160 votes; 95% CI 1192–1209 | Official leaderboard | Oct 2, 2026 |
| 75 | mistral-medium-3.5OW | Mistral AI | 1200 | 86 | Overall, style control on; 5379 votes; 95% CI 1190–1209 | Official leaderboard | Oct 2, 2026 |
| 76 | GPT-5.4 Nano | OpenAI | 1199 | 87 | Overall, style control on; 25647 votes; 95% CI 1193–1206“gpt-5.4-nano-high” | Official leaderboard | Oct 2, 2026 |
| 77 | Grok 4.1 fast | xAI | 1196 | 88 | Overall, style control on; 13385 votes; 95% CI 1189–1203“grok-4-1-fast-reasoning” | Official leaderboard | Oct 2, 2026 |
| 78 | Claude 3.7 Sonnet | Anthropic | 1193 | 89 | Overall, style control on; 1438 votes; 95% CI 1177–1208“claude-3-7-sonnet-20250219-thinking-32k” | Official leaderboard | Oct 2, 2026 |
| 79 | o1 | OpenAI | 1193 | 90 | Overall, style control on; 3694 votes; 95% CI 1182–1204“o1-2024-12-17” | Official leaderboard | Oct 2, 2026 |
| 80 | Gemini 2.5 Flash-Lite (Jun 2025) | 1188 | 92 | Overall, style control on; 35941 votes; 95% CI 1181–1196“gemini-2.5-flash-lite-preview-06-17-thinking” | Official leaderboard | Oct 2, 2026 | |
| 81 | qwen3-vl-235b-a22b-thinkingOW | Alibaba (Qwen) | 1188 | 94 | Overall, style control on; 2388 votes; 95% CI 1176–1200 | Official leaderboard | Oct 2, 2026 |
| 82 | qwen-vl-max-2025-08-13 | Alibaba (Qwen) | 1185 | 95 | Overall, style control on; 3173 votes; 95% CI 1173–1197 | Official leaderboard | Oct 2, 2026 |
| 83 | Grok 4 | xAI | 1184 | 96 | Overall, style control on; 30184 votes; 95% CI 1177–1192“grok-4-0709” | Official leaderboard | Oct 2, 2026 |
| 84 | GPT-5 mini | OpenAI | 1183 | 97 | Overall, style control on; 28965 votes; 95% CI 1175–1191“gpt-5-mini-high” | Official leaderboard | Oct 2, 2026 |
| 85 | Gemini 1.5 Pro (Sept 2024) | 1179 | 98 | Overall, style control on; 8902 votes; 95% CI 1170–1188“gemini-1.5-pro-002” | Official leaderboard | Oct 2, 2026 | |
| 86 | gemini-2.5-flash-lite-preview-09-2025-no-thinking | 1175 | 100 | Overall, style control on; 4862 votes; 95% CI 1165–1184 | Official leaderboard | Oct 2, 2026 | |
| 87 | Gemini 2.0 Flash (Feb 2025) | 1172 | 101 | Overall, style control on; 10134 votes; 95% CI 1165–1179“gemini-2.0-flash-001” | Official leaderboard | Oct 2, 2026 | |
| 88 | GPT-4o (May 2024) | OpenAI | 1162 | 102 | Overall, style control on; 23273 votes; 95% CI 1153–1171“gpt-4o-2024-05-13” | Official leaderboard | Oct 2, 2026 |
| 89 | glm-4.6vOW | Zhipu AI (Z.ai) | 1162 | 103 | Overall, style control on; 2455 votes; 95% CI 1148–1175 | Official leaderboard | Oct 2, 2026 |
| 90 | Claude 3.5 Sonnet | Anthropic | 1161 | 104 | Overall, style control on; 10374 votes; 95% CI 1153–1169“claude-3-5-sonnet-20241022” | Official leaderboard | Oct 2, 2026 |
| 91 | Gemma 3 27BOW | 1160 | 105 | Overall, style control on; 16989 votes; 95% CI 1152–1168“gemma-3-27b-it” | Official leaderboard | Oct 2, 2026 | |
| 92 | Mistral Medium 3.1 | Mistral AI | 1159 | 106 | Overall, style control on; 41639 votes; 95% CI 1153–1166“mistral-medium-2508” | Official leaderboard | Oct 2, 2026 |
| 93 | step-1o-turbo-202506 | StepFun | 1156 | 107 | Overall, style control on; 1897 votes; 95% CI 1141–1170 | Official leaderboard | Oct 2, 2026 |
| 94 | hunyuan-vision-1.5-thinking | Tencent | 1156 | 108 | Overall, style control on; 2429 votes; 95% CI 1143–1168 | Official leaderboard | Oct 2, 2026 |
| 95 | Mistral Medium 3 | Mistral AI | 1155 | 109 | Overall, style control on; 10809 votes; 95% CI 1147–1163“mistral-medium-2505” | Official leaderboard | Oct 2, 2026 |
| 96 | glm-4.5vOW | Zhipu AI (Z.ai) | 1153 | 110 | Overall, style control on; 3249 votes; 95% CI 1141–1164 | Official leaderboard | Oct 2, 2026 |
| 97 | hunyuan-large-vision | Tencent | 1149 | 111 | Overall, style control on; 1274 votes; 95% CI 1132–1165 | Official leaderboard | Oct 2, 2026 |
| 98 | GPT-5 nano | OpenAI | 1146 | 112 | Overall, style control on; 3930 votes; 95% CI 1135–1158“gpt-5-nano-high” | Official leaderboard | Oct 2, 2026 |
| 99 | Llama 4 MaverickOW | Meta | 1145 | 113 | Overall, style control on; 6923 votes; 95% CI 1136–1154“llama-4-maverick-17b-128e-instruct” | Official leaderboard | Oct 2, 2026 |
| 100 | step-3OW | StepFun | 1145 | 115 | Overall, style control on; 3196 votes; 95% CI 1133–1157 | Official leaderboard | Oct 2, 2026 |
Data sources & licences
| Source | Licence / terms | How we reuse it | Last fetched |
|---|---|---|---|
| LMArena leaderboard dataset LMArena, Arena Leaderboard Dataset (Hugging Face), CC BY 4.0 | CC BY 4.0 | Open licence – full data with attribution | Oct 10, 2026 ⚠︎ |
FAQ
What does LMArena Vision measure?
Arena for image-understanding chats: users upload images, compare two anonymous models and vote. Human preference on image + text questions.
How reliable is LMArena Vision?
Fewer votes than the text arena, so wider confidence intervals. Contamination risk: Low – Live prompts.