Atlas

Benchmarks

← All benchmarks

Arena Text — Hard Prompts English (No Style Control)

Chat & Writing

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

Top models (higher is better)

ModelScore
Claude Opus 5.51538
Gemini 4 Argon1534
Opus 4.61530
Claude Fable 5.11523
Claude Opus 51523
MiMo-V2.6-Pro1520
Claude Fable 51509
Opus 4.71508
Muse Wonder1506
Qwen3.8-Max1504
Gemini 3.8 Flash1504
Claude Sonnet 5.51502
Muse Spark 1.31501
Kimi K31497
GLM-5.31496
MiMo-V2.5-Pro1494
Sonnet 4.61493
MiMo-V2.6-Flash1492
Muse Spark 1.21489
Gemini 3.7 Flash1489
GLM-5.3 Flash1488
DeepSeek V4.1 Flash1488
Opus 4.81487
ERNIE 5.11484
Gemini 3.6 Flash1484
GPT-5.51483
Gemini 3.5 Flash1482
GLM-5.11482
Opus 4.51482
GLM-5.21481
Muse Spark 1.11480
Gemini 3.1 Pro Preview1480
GPT-5.41479
Muse Spark1477
GPT-5.6 Sol1477
Gemini 3 Pro Preview1475
MiMo-V2-Pro1473
Qwen3.7 Max Preview1473
Kimi K2.61471
Sonnet 4.51470
Loading Atlas data…