Sign in

Is Qwen3 235B A22B Thinking SOTA in Math?Not enough data#140 on LMArena Text (style-controlled, math); missing Humanity's Last Exam and FrontierMath

How it scores here

Qwen3 235B A22B Thinking in Math

Primary benchmarks

  • LMArena Text (style-controlled, math): 1397 (CI 1372–1422) on Nov 19, 2026 (#181 of 231).
Standings

Math

#ModelState
1GPT-6 AstraSOTA
2Claude Fable 5.1Near-SOTA
3Gemini 3.1 Pro PreviewNear-SOTA
4Claude Sonnet 5.5#4 in this category
5Gemini 3.8 Flash#5 in this category
6GPT-5.4 ProWas SOTA
7Muse SparkWas SOTA
8Gemini 3 Pro PreviewWas SOTA
9GPT-5.4#9 in this category
10Claude Opus 4.7#10 in this category
11Claude Opus 4.6#11 in this category
12GPT-5 ProWas SOTA

Category page →