Atlas

Benchmarks

← All benchmarks

Arena Text — Russian (No Style Control)

Chat & Writing

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

Top models (higher is better)

ModelScore
Gemini 4 Argon1537
Muse Wonder1522
Claude Fable 5.11518
Claude Opus 5.51517
Gemini 3.8 Flash1516
Gemini 3.7 Flash1514
Claude Opus 51508
Claude Fable 51503
Gemini 3.1 Pro Preview1497
Opus 4.61496
Opus 4.71494
Muse Spark 1.31494
Gemini 3.5 Flash1493
Gemini 3 Pro Preview1492
Gemini 3.6 Flash1488
Muse Spark 1.21484
Kimi K31483
Qwen3.7 Max Preview1483
Qwen3.8-Max1482
Muse Spark 1.11481
GPT-5.41480
Gemini 3 Flash Preview1480
MiMo-V2.6-Pro1479
Opus 4.81474
GPT-5.51473
GPT-5.6 Sol1472
GLM-5.31471
Muse Spark1467
DeepSeek V4.1 Flash1467
GLM-5.21467
GLM-5.3 Flash1466
Gemma 4 31B IT1461
Gemini 2.5 Pro1461
ERNIE 5.11459
GPT-6.1 Sol1458
Claude Sonnet 5.51458
Qwen3.7-Plus (source unspecified)1458
GLM-5.11455
DeepSeek-V4-Pro1453
GPT-5.6 Terra1453
Loading Atlas data…