Atlas

Benchmarks

← All benchmarks

LMArena Text Arena Instruction Following Elo

Chat & Writing · 2024-07-30

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 5.51521
Gemini 4 Argon1519
Opus 4.61511
Claude Fable 51507
Opus 4.71502
Claude Fable 5.11500
Muse Wonder1497
Claude Opus 51495
Opus 4.81491
GPT-5.6 Sol1491
Gemini 3.8 Flash1490
Kimi K31489
Gemini 3.7 Flash1487
Claude Sonnet 5.51485
GPT-6.1 Sol1485
Muse Spark 1.31485
Gemini 3.1 Pro Preview1480
GLM-5.31478
GPT-5.51478
DeepSeek V4.1 Flash1477
Qwen3.8-Max1475
Sonnet 4.61475
Opus 4.51474
GPT-6 Astra1473
GLM-5.3 Flash1473
Gemini 3.6 Flash1473
Gemini 3 Pro Preview1473
Muse Spark 1.21472
Muse Spark 1.11472
MiMo-V2.6-Pro1471
GPT-5.41471
GLM-5.21469
MiMo-V2.5-Pro1467
Sonnet 51466
Gemini 3.5 Flash1466
Qwen3.7 Max Preview1465
Muse Spark1465
GPT-5.6 Terra1464
GLM-5.11462
Sonnet 4.51462
Loading Atlas data…