Atlas

Benchmarks

← All benchmarks

LMArena Text Arena Elo

Chat & Writing · 2023-05-03

LMArena's Text Arena ranks language models from blind, pairwise user votes. Higher Elo ratings indicate stronger crowd preference.

Top models (higher is better)

ModelScore
Claude Fable 51509
Opus 4.61505
Opus 4.71502
Claude Opus 51492
Muse Spark 1.11490
Muse Spark1488
Gemini 3 Pro Preview1486
Gemini 3.1 Pro Preview1485
Kimi K31485
Opus 4.81484
GPT-5.6 Sol1483
Gemini 3.6 Flash1483
GPT-5.51482
GPT-5.41477
Gemini 3.5 Flash1476
Qwen3.7 Max Preview1475
Gemini 3 Flash Preview1473
Sonnet 4.61472
GLM-5.21469
Opus 4.51469
Grok 4.51469
GLM-5.11469
GPT-5.6 Terra1468
ERNIE 5.11468
Grok 4.11466
MiMo-V2.5-Pro1466
Sonnet 51462
Kimi K2.61461
Qwen3.7-Plus1459
DeepSeek-V4-Pro1458
GLM-51457
Gemini 3.5 Flash-Lite1457
Hy31456
Sonnet 4.51455
GPT-5.11455
Gemma 4 31B1451
GPT-5.6 Luna1450
Kimi K2.51450
GPT-5.3 Instant1449
GPT-5.4 Mini1449
Loading Atlas data…