Atlas

Benchmarks

← All benchmarks

Arena Vision — Chinese (No Style Control)

Multimodal

Raw Vision Arena rating for Chinese multimodal prompts without style-control adjustment. Higher ratings indicate stronger preference in blind pairwise user votes.

Top models (higher is better)

ModelScore
Gemini 3.8 Flash1416
Claude Opus 51400
Gemini 3.7 Flash1394
Claude Fable 5.11391
Muse Spark 1.31384
Gemini 3.5 Flash1377
Claude Fable 51373
Opus 4.71371
Gemini 3 Pro Preview1370
Opus 4.61370
Gemini 3.1 Pro Preview1364
GLM-5.3 Flash1363
GPT-6 Astra1361
GPT-5.51360
Qwen3.8-Max1358
Gemini 3.6 Flash1357
GPT-5.41354
Muse Spark 1.11348
Muse Spark1347
Kimi K2.61346
Opus 4.81346
Grok 4.51345
GPT-5.6 Terra1337
Qwen3.7-Plus (source unspecified)1332
Sonnet 51331
GLM 5V Turbo1329
Gemini 3 Flash Preview1329
DeepSeek V4.1 Flash1327
Gemma 4 31B IT1326
MiMo-V2.51326
Qwen3.8 27B1326
Kimi K2.51324
Qwen3.5 397B A17B1314
GPT-5.6 Sol1313
Gemma 4 26B A4B IT1312
MiniMax M31311
Gemini 3.5 Flash-Lite1311
Sonnet 4.61307
GPT-5.21305
Qwen3.5 122B A10B1301
Loading Atlas data…