Atlas

Benchmarks

← All benchmarks

LMArena Text Arena Chinese Elo

Chat & Writing · 2024-04-14

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 51590
Claude Fable 51556
Opus 4.61555
Opus 4.71544
GPT-5.6 Sol1542
Muse Spark 1.11541
Gemini 3.6 Flash1535
Gemini 3.1 Pro Preview1530
GPT-5.51530
Kimi K31529
Kimi K2.61529
Opus 4.81529
Gemini 3 Pro Preview1527
Grok 4.51525
GLM-5.21523
GLM-51523
GPT-5.6 Terra1522
Muse Spark1522
Qwen3.7 Max Preview1522
Gemini 3 Flash Preview1520
Hy31520
GLM-5.11517
Gemini 3.5 Flash1515
MiMo-V2.5-Pro1511
Sonnet 4.61508
GPT-5.41505
Qwen3.7-Plus1505
Kimi K2.51503
Grok 4.11503
Qwen3.5 397B A17B1501
DeepSeek-V4-Pro1500
ERNIE 5.11500
Opus 4.51499
Sonnet 51497
Gemini 3.5 Flash-Lite1496
GPT-5.11495
GPT-5.3 Instant1490
Gemini 2.5 Pro1490
GPT-5.6 Luna1489
Qwen3 Max Preview1488
Loading Atlas data…