Benchmarks
Reasoning benchmarks
We track 227 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 101–200 of 227 benchmarks.
- Natural2Code8 models
- OJBench8 models
- Wild Bench8 models
- ZebraLogic8 models
- Graphwalks parents >128k7 models
- MRCR7 models
- V*7 models
- AA-Omniscience Index6 models
- AutomationBench-AA6 models
- ComplexFuncBench6 models
- Chartography (no tools)5 models
- GeoBench AVW5 models
- GeoBench Rural5 models
- GeoBench Urban5 models
- OmniScience5 models
- AetherCode4 models
- Minerva4 models
- OneMillion Bench4 models
- PerceptionBench4 models
- ProfBench4 models
- Social IQa4 models
- TempCompass4 models
- TIR-Bench4 models
- TOMATO4 models
- VisuLogic4 models
- Workspace Bench4 models
- ARC-E3 models
- CommonSenseQA3 models
- ContPhy3 models
- CRAG3 models
- CrossVid3 models
- FActScore3 models
- FRAMES3 models
- GraphWalks3 models
- Graphwalks BFS 1M3 models
- HLE-Verified3 models
- MRCR v23 models
- Natural Questions3 models
- Nuscene3 models
- SlakeVQA3 models
- TreeBench3 models
- AA Intelligence Index v4.32 models
- AITZ_EM2 models
- Android Control High_EM2 models
- Android Control Low_EM2 models
- AutoLogi2 models
- BFCL_v3_MultiTurn2 models
- CLUEWSC2 models
- CreativeWork2 models
- FinSearchComp-T32 models
- FlenQA2 models
- GameWorld2 models
- IPhO 20252 models
- KINA2 models
- LongFact Concepts2 models
- LongFact Objects2 models
- MeasureBench2 models
- MRCR 1M2 models
- MSQA2 models
- PerceptionTest2 models
- SuperGLUE2 models
- xDailyBench2 models
- AdvancedIF1 model
- Arc1 model
- Biorisk knowledge1 model
- BixBench MCQ1 model
- CloningScenarios1 model
- Cohere Data Analysis1 model
- Cohere Memory Usage Quality1 model
- CorpusQA1 model