Atlas

Benchmarks

← All benchmarks

Arena Text — Occupational Mathematical (No Style Control)

Math

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

Top models (higher is better)

ModelScore
Gemini 4 Argon1549
Claude Opus 51543
Claude Fable 5.11543
Opus 4.61525
Claude Opus 5.51525
Gemini 3.8 Flash1520
Claude Fable 51515
GPT-6.1 Sol1514
MiMo-V2.6-Pro1511
Claude Sonnet 5.51511
Muse Wonder1511
Opus 4.71509
Kimi K31504
Gemini 3.7 Flash1502
GLM-5.31498
GPT-5.41497
Opus 4.81497
GLM-5.3 Flash1497
GPT-5.51496
GPT-5.6 Sol1495
Gemini 3.5 Flash1494
Muse Spark 1.31493
Gemini 3.6 Flash1492
Qwen3.8-Max1491
ERNIE 5.11491
MiMo-V2.5-Pro1491
GLM-5.21490
Muse Spark 1.11485
DeepSeek V4.1 Flash1484
Hy31483
Sonnet 4.61483
GPT-6 Astra1482
Sonnet 51482
Grok 4.51482
Muse Spark 1.21481
Kimi K2.61480
Gemini 3.1 Pro Preview1478
MiMo-V2.6-Flash1478
GPT-5.6 Terra1478
Gemma 4 26B A4B IT1476
Loading Atlas data…