now jobs Jobs market snapshot refreshed now data Recomputed internal benchmark aggregates now data Attempted Chatbot Arena collection now data Updated speed measurements now data Pulled latest OpenRouter price index now data Validated official pricing snapshots 17h digest Meta has some Wonder Full new AI models 17h digest Meta releases its first frontier model, Muse Wonder 17h digest Published the 2026-10-12 daily digest 17h digest WindBorne wins a hefty US Air Force contract for its balloon data now jobs Jobs market snapshot refreshed now data Recomputed internal benchmark aggregates now data Attempted Chatbot Arena collection now data Updated speed measurements now data Pulled latest OpenRouter price index now data Validated official pricing snapshots 17h digest Meta has some Wonder Full new AI models 17h digest Meta releases its first frontier model, Muse Wonder 17h digest Published the 2026-10-12 daily digest 17h digest WindBorne wins a hefty US Air Force contract for its balloon data

nowJobs market snapshot refreshed nowRecomputed internal benchmark aggregates nowAttempted Chatbot Arena collection nowUpdated speed measurements nowPulled latest OpenRouter price index nowValidated official pricing snapshots 17hMeta has some Wonder Full new AI models 17hMeta releases its first frontier model, Muse Wonder 17hPublished the 2026-10-12 daily digest 17hWindBorne wins a hefty US Air Force contract for its balloon data nowJobs market snapshot refreshed nowRecomputed internal benchmark aggregates nowAttempted Chatbot Arena collection nowUpdated speed measurements nowPulled latest OpenRouter price index nowValidated official pricing snapshots 17hMeta has some Wonder Full new AI models 17hMeta releases its first frontier model, Muse Wonder 17hPublished the 2026-10-12 daily digest 17hWindBorne wins a hefty US Air Force contract for its balloon data

Leaderboard

Best AI for Math

Models ranked by mathematical reasoning performance across MATH-500, GSM8K, and AIME 2025. Composite score averages available math benchmarks.

#	Model	Provider	Score	MATH-500	AIME 2025	GSM8K	Input $/M	Open Source

Composite score averages only current, row-level source-backed MATH-500, GSM8K and AIME 2025 results. Models without rankable maths evidence are not ranked.