Atlas

Benchmarks

← All benchmarks

Arena Document (Style Controlled)

Professional Work

Style-controlled Document Arena rating for models analyzing document and long-content inputs. Higher ratings indicate stronger preference in blind pairwise user votes after adjusting for response style.

Top models (higher is better)

ModelScore
Claude Fable 51499
Opus 4.61495
Opus 4.71494
Claude Fable 5.11493
Claude Opus 51490
GPT-5.6 Sol1489
Opus 4.81486
GPT-5.51483
Muse Spark 1.11479
Sonnet 4.61478
GPT-5.6 Terra1477
Sonnet 51476
GPT-6 Astra1473
GPT-5.41470
Muse Spark 1.31468
Muse Spark1467
Opus 4.51465
Grok 4.51464
GPT-5.6 Luna1462
Grok 4.61461
Gemini 3.5 Flash1461
Gemini 3.1 Pro Preview1459
Gemini 3.6 Flash1453
Qwen3.7-Plus (source unspecified)1453
Gemini 3 Pro Preview1451
Kimi K2.61450
Sonnet 4.51449
Gemma 4 31B IT1445
Gemini 3 Flash Preview1439
Kimi K2.51436
MiniMax M31433
Gemini 2.5 Pro1431
GPT-5.21425
Haiku 4.51419
GPT-5.11416
GLM 5V Turbo1400
Loading Atlas data…