Atlas

Benchmarks

← All benchmarks

Arena Text — Software & IT Services (No Style Control)

Professional Work

Raw Arena rating for software and IT-services Text Arena prompts without style-control adjustment. Higher ratings indicate stronger preference in blind pairwise user votes.

Top models (higher is better)

ModelScore
Claude Opus 51528
Opus 4.61528
Claude Fable 51516
Opus 4.71515
Muse Spark 1.11499
Qwen3.7 Max Preview1498
Gemini 3.6 Flash1493
Gemini 3.5 Flash1492
GPT-5.51492
ERNIE 5.11492
MiMo-V2.5-Pro1491
Kimi K31491
GPT-5.41491
Sonnet 4.61490
GLM-5.11488
Gemini 3 Pro Preview1487
Gemini 3.1 Pro Preview1486
Opus 4.81486
Muse Spark1484
Kimi K2.61484
GPT-5.6 Sol1483
GPT-5.6 Terra1481
Opus 4.51481
Sonnet 51481
Grok 4.51478
Qwen3.7-Plus1478
GLM-5.21477
Sonnet 4.51472
Gemini 3 Flash Preview1471
DeepSeek-V4-Pro1470
Kimi K2.51468
MiniMax M31468
MiMo-V2-Pro1468
Inkling1465
Hy31464
GPT-5.6 Luna1464
MiMo-V2.51463
Qwen3.5 397B A17B1462
Qwen3.6 Plus (2026-04-02)1461
Gemma 4 31B1461
Loading Atlas data…