Atlas

Benchmarks

← All benchmarks

LMArena Text Arena Hard Prompts Elo

Chat & Writing · 2024-05-17

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 Fable 51534
Opus 4.61534
Opus 4.71525
Claude Opus 51520
Opus 4.81513
Muse Spark 1.11511
Gemini 3.1 Pro Preview1507
Kimi K31506
GPT-5.6 Sol1505
Muse Spark1505
Gemini 3 Pro Preview1504
Sonnet 4.61503
Gemini 3.6 Flash1502
GPT-5.51502
GPT-5.41499
Opus 4.51498
Qwen3.7 Max Preview1495
MiMo-V2.5-Pro1494
Sonnet 51493
Gemini 3.5 Flash1493
Gemini 3 Flash Preview1492
GLM-5.11491
Grok 4.51491
GPT-5.6 Terra1489
GLM-5.21488
ERNIE 5.11487
Kimi K2.61485
Sonnet 4.51483
DeepSeek-V4-Pro1481
Qwen3.7-Plus1478
GLM-51478
Grok 4.11478
Opus 4.11477
MiMo-V2-Pro1476
Gemini 3.5 Flash-Lite1476
Hy31475
GPT-5.6 Luna1474
GPT-5.11474
Gemma 4 31B1473
Kimi K2.51472
Loading Atlas data…