Atlas

Benchmarks

← All benchmarks

Arena Text — Japanese (No Style Control)

Chat & Writing

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

Top models (higher is better)

ModelScore
Claude Fable 5.11560
Claude Opus 5.51535
Claude Opus 51521
Gemini 3.7 Flash1513
Gemini 3.8 Flash1511
Gemini 3 Pro Preview1509
Claude Fable 51506
GPT-5.51499
Kimi K31497
Gemini 3.1 Pro Preview1493
Gemini 3 Flash Preview1490
Opus 4.61488
Gemini 3.5 Flash1487
GPT-5.41485
Gemini 3.6 Flash1475
Opus 4.71474
GPT-5.6 Sol1473
Qwen3.8-Max1473
Muse Spark 1.31465
Gemini 2.5 Pro1461
Claude Sonnet 5.51460
GLM-5.31458
GPT-5.11454
GPT-5.6 Terra1453
Muse Spark 1.11450
GLM-5.21449
DeepSeek-V4-Pro1446
Kimi K2.61443
Opus 4.81442
GLM-5.3 Flash1442
GLM-4.71439
GLM-5.11435
Grok 4.51431
Qwen3.5 397B A17B1430
Inkling1428
DeepSeek V4.1 Flash1428
Gemini 3.5 Flash-Lite1426
Sonnet 51423
Kimi K2.51422
ERNIE 5.11422
Loading Atlas data…