LMArena Vision: Chinese leaderboard
- models measured
- 72
- lead over #2
- +16 Elo
- median
- 1303
Leader on LMArena Vision: Chinese · read October 21
On October 21, 2026, 72 of our catalog's models had a LMArena Vision: Chinese result, and Google Gemini 3.8 Flash led with 1416.
Image prompts written in Chinese. People ask a question, compare two anonymous answers and vote for the better one; the rating is built from those votes, as LMArena publishes it.
Role in the index
We collect this benchmark and show it as supporting evidence. It is not part of the Vibe Coding Index, because it measures something other than coding and the index is about coding. Read the methodology
| # | Model | Result | Votes | Interval (±) | Evidence | Coding index | Output $/Mtok |
|---|---|---|---|---|---|---|---|
| 1 | Gemini 3.8 FlashGoogle | 1416 | 224 | ±39.3 | 70.5 | $4.125 | |
| 2 | Claude Opus 5Anthropic | 1400 | 1,037 | ±20.2 | 81.8 | $25.00 | |
| 3 | Gemini 3.7 FlashGoogle | 1394 | 302 | ±34.1 | 69.9 | $3.75 | |
| 4 | Claude Fable 5.1Anthropic | 1391 | 284 | ±34.7 | 89.1 | $50.00 | |
| 5 | Muse Spark 1.3Meta | 1384 | 257 | ±36.2 | 79.7 | $4.25 | |
| 6 | Gemini 3.5 FlashGoogle | 1377 | 847 | ±22.3 | 42.8 | $9.00 | |
| 7 | Claude Fable 5Anthropic | 1373 | 784 | ±22.7 | 82.1 | $50.00 | |
| 8 | Claude Opus 4.7Anthropic | 1371 | 1,504 | ±17.6 | 51.1 | $25.00 | |
| 9 | Claude Opus 4.6Anthropic | 1370 | 1,681 | ±17.1 | 66.9 | $25.00 | |
| 10 | Gemini 3.1 Pro PreviewGoogle | 1364 | 3,186 | ±13.3 | 35.3 | $12.00 | |
| 11 | GLM 5.3 FlashZ.ai | 1363 | 699 | ±23.3 | 71.0 | $0.50 | |
| 12 | GPT-6 AstraOpenAI | 1361 | 259 | ±37.0 | 88.7 | $50.00 | |
| 13 | GPT-5.5OpenAI | 1360 | 1,777 | ±16.6 | 67.8 | $30.00 | |
| 14 | Qwen3.8 Max (0902)Qwen | 1358 | 541 | ±26.0 | 63.5 | $6.00 | |
| 15 | Gemini 3.6 FlashGoogle | 1357 | 536 | ±26.1 | 48.3 | $3.75 | |
| 16 | GPT-5.4OpenAI | 1354 | 1,846 | ±16.2 | 37.7 | $15.00 | |
| 17 | Muse Spark 1.1Meta | 1348 | 767 | ±22.4 | 48.9 | $4.25 | |
| 18 | Kimi K2.6Moonshot AI | 1346 | 1,027 | ±20.4 | 44.2 | $3.41 | |
| 19 | Claude Opus 4.8Anthropic | 1346 | 1,131 | ±19.6 | 66.0 | $25.00 | |
| 20 | Grok 4.5xAI | 1345 | 728 | ±23.3 | 75.5 | $6.00 | |
| 21 | GPT-5.6 TerraOpenAI | 1337 | 682 | ±23.6 | 69.0 | $12.00 | |
| 22 | Qwen3.7 PlusQwen | 1332 | 740 | ±22.8 | Not benchmarked | $1.28 | |
| 23 | Claude Sonnet 5Anthropic | 1331 | 847 | ±21.9 | 63.5 | $10.00 | |
| 24 | GLM 5V TurboZ.ai | 1329 | 2,644 | ±14.3 | 28.8 | $4.00 | |
| 25 | Kimi K2.5Moonshot AI | 1328 | 273 | ±34.0 | 58.5 | $2.50 | |
| 26 | DeepSeek V4.1 FlashDeepSeek | 1327 | 199 | ±41.5 | 77.2 | $0.525 | |
| 27 | Gemma 4 31BGoogle | 1326 | 2,734 | ±14.1 | 38.3 | $0.40 | |
| 28 | MiMo-V2.5Xiaomi | 1326 | 1,580 | ±17.0 | 34.0 | $0.238 | |
| 29 | Qwen3.8 27BQwen | 1326 | 409 | ±29.8 | 73.6 | $1.35 | |
| 30 | Qwen3.5 397B A17BQwen | 1314 | 2,045 | ±15.3 | 28.6 | $2.34 | |
| 31 | GPT-5.6 SolOpenAI | 1313 | 677 | ±23.9 | 76.7 | $20.00 | |
| 32 | Gemma 4 26B A4BGoogle | 1312 | 1,609 | ±17.0 | 22.8 | $0.255 | |
| 33 | MiniMax M3MiniMax | 1311 | 1,048 | ±20.2 | 57.8 | $1.20 | |
| 34 | Gemini 3.5 Flash LiteGoogle | 1311 | 524 | ±26.7 | 33.6 | $2.50 | |
| 35 | Claude Sonnet 4.6Anthropic | 1307 | 1,736 | ±16.8 | 57.8 | $15.00 | |
| 36 | GPT-5.2OpenAI | 1305 | 1,351 | ±17.9 | 58.9 | $14.00 | |
| 37 | Qwen3.5-122B-A10BQwen | 1301 | 1,011 | ±20.3 | 22.9 | $2.08 | |
| 38 | Qwen3.5-27BQwen | 1301 | 1,997 | ±15.5 | 22.6 | $2.16 | |
| 39 | Gemini 2.5 ProGoogle | 1300 | 4,797 | ±12.9 | 37.0 | $10.00 | |
| 40 | GPT-5.6 LunaOpenAI | 1294 | 631 | ±24.9 | 61.5 | $1.20 | |
| 41 | T Inkling SmallThinking Machines | 1292 | 461 | ±28.4 | 29.9 | $1.20 | |
| 42 | Grok 4.6xAI | 1289 | 380 | ±30.7 | 76.4 | $6.00 | |
| 43 | GPT-5.4 MiniOpenAI | 1286 | 1,661 | ±17.1 | 28.2 | $4.50 | |
| 44 | Gemini 2.5 FlashGoogle | 1284 | 3,692 | ±13.0 | 28.7 | $2.50 | |
| 45 | Qwen3 VL 235B A22B InstructQwen | 1283 | 998 | ±19.4 | Not benchmarked | $1.90 | |
| 46 | Gemini 3.1 Flash Lite PreviewGoogle | 1278 | 1,989 | ±15.8 | 8.1 | $1.50 | |
| 47 | GPT-5.1OpenAI | 1270 | 671 | ±23.3 | 53.3 | $10.00 | |
| 48 | GPT-5OpenAI | 1267 | 2,052 | ±16.8 | 72.2 | $10.00 | |
| 49 | GPT-6 LunaOpenAI | 1266 | 169 | ±43.9 | 71.4 | $0.50 | |
| 50 | Grok 4.3xAI | 1257 | 1,650 | ±17.2 | 22.5 | $2.50 | |
| 51 | Grok 4.7xAI | 1250 | 182 | ±45.0 | 75.0 | $6.00 | |
| 52 | Mistral Medium 3.5Mistral AI | 1244 | 287 | ±37.5 | 9.2 | $7.50 | |
| 53 | o3OpenAI | 1237 | 2,258 | ±16.3 | 70.7 | $8.00 | |
| 54 | o1OpenAI | 1235 | 99 | ±58.2 | 61.7 | $60.00 | |
| 55 | GPT-5.4 NanoOpenAI | 1232 | 1,739 | ±17.0 | Not benchmarked | $1.25 | |
| 56 | GPT-4.1OpenAI | 1226 | 1,978 | ±16.8 | 46.5 | $8.00 | |
| 57 | o4 MiniOpenAI | 1218 | 1,934 | ±17.2 | 68.6 | $4.40 | |
| 58 | GLM 4.6VZ.ai | 1217 | 166 | ±41.2 | Not benchmarked | $0.90 | |
| 59 | GPT-5 MiniOpenAI | 1214 | 1,600 | ±18.5 | 59.8 | $2.00 | |
| 60 | Mistral Medium 3.1Mistral AI | 1190 | 2,176 | ±16.1 | Not benchmarked | $2.00 | |
| 61 | GPT-4.1 MiniOpenAI | 1186 | 1,778 | ±17.7 | 28.5 | $1.60 | |
| 62 | GPT-5 NanoOpenAI | 1176 | 189 | ±43.2 | 34.8 | $0.40 | |
| 63 | Gemma 3 27BGoogle | 1164 | 1,020 | ±20.9 | 4.9 | $0.35 | |
| 64 | Mistral Medium 3Mistral AI | 1161 | 599 | ±25.1 | Not benchmarked | $2.00 | |
| 65 | Mistral Small 3.1 24BMistral AI | 1155 | 1,251 | ±20.7 | Not benchmarked | $0.555 | |
| 66 | Llama 4 MaverickMeta Llama | 1141 | 294 | ±34.1 | 18.1 | $0.6525 | |
| 67 | GPT-4o (2024-05-13)OpenAI | 1132 | 1,472 | ±38.2 | Not benchmarked | $15.00 | |
| 68 | Qwen2.5 VL 72B InstructQwen | 1107 | 119 | ±55.3 | Not benchmarked | $1.00 | |
| 69 | Llama 4 ScoutMeta Llama | 1106 | 278 | ±36.3 | 9.1 | $0.30 | |
| 70 | GPT-4 TurboOpenAI | 1045 | 937 | ±40.7 | Not benchmarked | $30.00 | |
| 71 | GPT-4o-mini (2024-07-18)OpenAI | 1033 | 954 | ±38.1 | 3.6 | $0.60 | |
| 72 | GPT-4o (2024-08-06)OpenAI | 1009 | 186 | ±52.2 | 23.1 | $10.00 |
Only models in our catalog are listed, and each appears once with its best published run. 172 rows were read from the source on October 21, 2026, and rows that match no catalog model are left out. Every other column is a figure the board publishes for the same run.