Atlas

Benchmarks

← All benchmarks

Arena Vision — Chinese (Style Controlled)

Multimodal

Style-controlled Vision Arena rating for Chinese multimodal prompts. Higher ratings indicate stronger preference in blind pairwise user votes after adjusting for response style.

Top models (higher is better)

ModelScore
Gemini 3.5 Flash1372
Claude Fable 51369
Muse Spark 1.11363
Opus 4.61360
Gemini 3 Pro Preview1352
Gemini 3.1 Pro Preview1345
Sonnet 51344
Opus 4.71344
Muse Spark1341
Opus 4.81335
Kimi K2.61329
Gemini 3 Flash Preview1323
GPT-5.51320
Qwen3.7-Plus1317
MiMo-V2.51316
GPT-5.41316
Gemma 4 31B1313
MiniMax M31303
Sonnet 4.61302
Qwen3.5 397B A17B1298
Kimi K2.51297
GPT-5.4 Mini1295
Gemma 4 26B A4B1295
GLM 5V Turbo1292
GPT-5.21291
Grok 4.31284
Qwen3.5 122B A10B1278
Gemini 3.1 Flash-Lite Preview1276
GPT-5.11276
MiMo-V2-Omni1270
Gemini 2.5 Pro1267
Gemini 2.5 Flash Preview (09-2025)1258
Gemini 2.5 Flash1256
GPT-4.1 (2025-04-14)1240
GPT-5.4 Nano1238
GPT-51227
GPT-4.1 Mini (2025-04-14)1226
O3 (2025-04-16)1225
Grok 4.1 Fast1222
Grok 41213
Loading Atlas data…