Atlas

Benchmarks

← All benchmarks

Arena Text — Creative Writing (No Style Control)

Chat & Writing

Raw Arena rating for creative-writing Text Arena prompts without style-control adjustment. Higher ratings indicate stronger preference in blind pairwise user votes.

Top models (higher is better)

ModelScore
Opus 4.61505
Claude Fable 51499
Claude Opus 51498
Opus 4.71483
Gemini 3 Pro Preview1482
Gemini 3.1 Pro Preview1481
Gemini 3.5 Flash1470
Kimi K31461
Gemini 3 Flash Preview1458
Muse Spark1458
GLM-5.21456
Gemini 3.6 Flash1455
Gemini 2.5 Pro1454
Qwen3.7 Max Preview1453
GPT-5.6 Sol1453
GPT-5.51453
GLM-5.11453
Opus 4.81449
DeepSeek-V4-Pro1446
Qwen3.7-Plus1444
Sonnet 4.51441
Opus 4.51441
ERNIE 5.11440
GPT-5.41439
Muse Spark 1.11439
Grok 4.51438
Sonnet 4.61438
GLM-51436
MiMo-V2.5-Pro1434
Kimi K2.61430
GPT-5.11428
Kimi K2.51423
Gemini 3.5 Flash-Lite1418
Gemma 4 31B1417
MiMo-V2-Pro1416
GLM 4.61413
Grok 4.11412
Opus 4.11410
Sonnet 51410
DeepSeek-R1-05281410
Loading Atlas data…