Atlas

Benchmarks

← All benchmarks

Arena Text — Exclude Ties (Style Controlled)

Chat & Writing

Style-controlled Text Arena rating computed with tied votes excluded. It is separate from the overall series because removing ties changes the vote sample and rating estimates.

Top models (higher is better)

ModelScore
Gemini 4 Argon1543
Muse Wonder1533
Claude Opus 5.51530
Claude Fable 51523
Claude Fable 5.11522
Opus 4.61522
Opus 4.71519
Muse Spark 1.31514
Gemini 3.8 Flash1514
Muse Spark 1.21507
Claude Opus 51506
Gemini 3.7 Flash1505
Muse Spark 1.11505
Muse Spark1502
Gemini 3.1 Pro Preview1502
Kimi K31501
Gemini 3 Pro Preview1501
GPT-5.6 Sol1501
Qwen3.8-Max1496
Gemini 3.6 Flash1495
Opus 4.81494
GPT-6.1 Sol1493
GPT-5.51492
GLM-5.31490
MiMo-V2.6-Pro1489
GPT-6 Astra1489
Gemini 3.5 Flash1487
Claude Sonnet 5.51487
DeepSeek V4.1 Flash1485
GLM-5.21485
Gemini 3 Flash Preview1485
GPT-5.41485
GLM-5.3 Flash1484
Sonnet 4.61482
Qwen3.7 Max Preview1482
Opus 4.51479
GPT-5.6 Terra1475
ERNIE 5.11473
Grok 4.11473
Grok 4.51473
Loading Atlas data…