Atlas

Benchmarks

← All benchmarks

Arena Vision — English (No Style Control)

Multimodal

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

Top models (higher is better)

ModelScore
Claude Fable 51336
Gemini 3.6 Flash1330
Opus 4.61322
Opus 4.71320
Claude Opus 51320
Muse Spark 1.11307
Muse Spark1304
Gemini 3.5 Flash1300
Sonnet 4.61298
GPT-5.41297
Opus 4.81294
GPT-5.51294
Gemini 3 Pro Preview1293
Grok 4.51291
Gemini 3.1 Pro Preview1289
Kimi K2.61286
Sonnet 51278
GLM 5V Turbo1275
Qwen3.5 397B A17B1273
Gemini 3 Flash Preview1272
Qwen3.7-Plus1271
Kimi K2.51270
GPT-5.6 Sol1270
Gemma 4 31B1268
MiniMax M31268
Gemini 3.5 Flash-Lite1266
MiMo-V2.51266
GPT-5.6 Terra1264
Gemini 2.5 Pro1262
Gemini 2.5 Flash Preview (09-2025)1255
GPT-5.6 Luna1253
Gemma 4 26B A4B1253
GPT-5.11252
Qwen3.5 122B A10B1247
GPT-5.21245
GPT-5.4 Mini1243
Gemini 2.5 Flash1242
MiMo-V2-Omni1241
Grok 4.31237
Gemini 3.1 Flash-Lite Preview1229
Loading Atlas data…