Benchmarks
Math benchmarks
These benchmarks measure how well models solve maths problems.
We track 147 benchmarks in this topic. Main benchmarks are listed by how many models they cover, most first. Where a benchmark has separate results for a subset of its tasks, those results appear right under it. Click one for its leaderboard.
Showing 101–147 of 147 benchmarks.
- TheoremQA6 models
- AlignBench5 models
- HMMT February 20265 models
- EMMA4 models
- MathArena Putnam 20254 models
- AMC_2022_233 models
- BIG-Bench3 models
- FinQA3 models
- MATH (CoT)3 models
- Multilingual MGSM (CoT)3 models
- PhiBench3 models
- AIME 2026, part of AIME40 models
- CNMO 20242 models
- FunctionalMATH2 models
- GSM-8K (CoT)2 models
- IMO 20252 models
- InterGPS2 models
- LiveMathematicianBench2 models
- MathVerse2 models
- PhysicsFinals2 models
- AMO Bench1 model
- CBNSL1 model
- GSM8K Chat1 model
- IMOProof-Adv1 model
- MathArena IMC 20251 model
- MathVerse-Mini1 model
- MT-AIME 20251 model
- OlympiadBench1 model
- PolyMath-en1 model
- SAT Math1 model
- STEM1 model
- USAMO251 model