Atlas

Benchmarks

← All benchmarks

LMArena Text Overall No Style Control Elo

Chat & Writing · 2024-08-26

Category-specific Elo score from the LMArena text-arena voting leaderboard. Higher Elo indicates stronger crowd-voted preference within this category.

Top models (higher is better)

ModelScore
Claude Opus 51507
Opus 4.61503
Claude Fable 51496
Opus 4.71489
Gemini 3.5 Flash1480
Gemini 3 Pro Preview1480
Gemini 3.1 Pro Preview1479
Gemini 3.6 Flash1478
Muse Spark 1.11478
Qwen3.7 Max Preview1475
Muse Spark1473
Kimi K31473
GPT-5.51470
GPT-5.41470
ERNIE 5.11468
Gemini 3 Flash Preview1466
GLM-5.11465
GLM-5.21464
Opus 4.81463
MiMo-V2.5-Pro1462
Qwen3.7-Plus1459
Gemini 2.5 Pro1458
Sonnet 4.61457
Kimi K2.61455
GPT-5.6 Sol1453
Grok 4.51452
DeepSeek-V4-Pro1451
Opus 4.51449
GPT-5.6 Terra1448
Kimi K2.51445
GLM-51445
Sonnet 51443
Gemma 4 31B1441
Hy31441
GPT-5.11441
GLM 4.61440
Inkling1440
Qwen3 Max Preview1439
Qwen3.5 397B A17B1438
Sonnet 4.51438
Loading Atlas data…