Atlas

Benchmarks

← All benchmarks

ArXivMath 08/2026 (Native Harnesses)

Math · 2026-09-15

MathArena's August 2026 research-mathematics release: 57 questions, native model harnesses, Python and SageMath without internet, and Gemini 3.8 Flash answer-equivalence grading. New construction emphasizes resolved conjectures; this tool-enabled protocol is distinct from earlier monthly evaluations. Time and cost limits were introduced partway through the published runs.

Top models (higher is better)

ModelScore
GPT-6 Astra88.6
Claude Fable 5.187.7
Muse Spark 1.345.6
Gemini 3.8 Flash40.4
Kimi K338.6
Loading Atlas data…