Atlas

Benchmarks

← All benchmarks

Arena Text — Non-English (No Style Control)

Chat & Writing

Raw Arena rating for the non-English Text Arena aggregate without style-control adjustment. Higher ratings indicate stronger preference in blind pairwise user votes.

Top models (higher is better)

ModelScore
Gemini 4 Argon1519
Claude Fable 5.11508
Claude Opus 5.51506
Claude Opus 51501
Muse Wonder1491
Gemini 3.8 Flash1491
Opus 4.61488
Gemini 3.7 Flash1487
Muse Spark 1.31482
Claude Fable 51481
Opus 4.71480
Gemini 3.1 Pro Preview1476
Gemini 3.5 Flash1475
Muse Spark 1.21475
Gemini 3 Pro Preview1474
MiMo-V2.6-Pro1474
Qwen3.7 Max Preview1473
Qwen3.8-Max1473
Muse Spark 1.11471
Gemini 3.6 Flash1470
GPT-5.51467
GPT-5.41465
Kimi K31465
Muse Spark1464
GLM-5.3 Flash1463
GLM-5.31460
GLM-5.21458
Gemini 3 Flash Preview1458
Claude Sonnet 5.51456
ERNIE 5.11454
Gemini 2.5 Pro1452
GPT-5.6 Sol1451
Opus 4.81450
DeepSeek V4.1 Flash1450
GLM-5.11449
MiMo-V2.5-Pro1448
Kimi K2.61446
Qwen3.7-Plus (source unspecified)1445
GPT-5.6 Terra1441
Sonnet 4.61440
Loading Atlas data…