Registry benchmarks

Registry benchmarks
BenchmarkLatest versionReleased
Agent's Last ExamALE-V1 — ALE-Claw, 5h (configuration)September 30, 2026
AI2DTESTMarch 24, 2016
APEX-AgentsOriginalJanuary 21, 2026
Arena TextText Overall — 2026-09-13 snapshotSeptember 13, 2026
AutomationBench1.0.6July 31, 2026
BabyVisionOriginalJanuary 13, 2026
BenchCADVision2Code — With tools (configuration)June 24, 2026
BeyondAIMEOriginalJune 17, 2025 at 3:37 AM UTC
BIG-Bench Extra HardOriginal — micro averageFebruary 26, 2025
BrowseCompBrowseCompApril 10, 2025
ChartographyOctober 2026 leaderboardOctober 7, 2026
ChartQATESTMarch 19, 2022
CharXivReasoning — Search and code execution (configuration)July 21, 2026
CL-benchOriginalFebruary 3, 2026
CodeforcesNo tools (configuration)August 5, 2025
Coronavirus–ACE2 Cell-Entry ScreenOriginalSeptember 3, 2026
CursorBench4.0September 10, 2026
CWE-bench1September 25, 2026
CyberGymOriginalJune 3, 2025
DeepSearchQA900-prompt setDecember 11, 2025
DeepSWE1.1June 15, 2026
ERQANo tools (configuration)April 8, 2026
Finance Agent Benchmark2 — October 2026 evaluationOctober 6, 2026
FrontierScienceResearchDecember 16, 2025
FrontierSWE2 — September 2026 evaluationSeptember 30, 2026
GAIA2Public 800 tasksFebruary 12, 2026
GDP.pdfGDP.pdfJuly 13, 2026
GDPvalGDPvalSeptember 25, 2025
Global MMLUOriginalDecember 4, 2024
GPQADiamondNovember 20, 2023
GraphWalks256k to 1M — BFS, 200 itemsSeptember 30, 2026
HallusionBenchOriginalOctober 23, 2023
HLABOctober 2026 evaluationOctober 6, 2026
HealthBenchHealthBenchMay 12, 2025
HLE-VerifiedFull 1,811-item verified setFebruary 15, 2026
Humanity's Last ExamFull set — Scale AI evaluationApril 3, 2025
IFBenchOriginalJuly 3, 2025
IFEvalOriginalNovember 14, 2023
International Physics Olympiad2025 TheoryJuly 21, 2025
JobBenchMain 65 tasksMay 25, 2026
LABBench2Terminal and internet — September 2026 evaluation (configuration)September 30, 2026
LongBenchv2December 20, 2024
LVBench1024 frames — AgenticJune 12, 2024
MMMUMMMUNovember 27, 2023
MMLUMMLUSeptember 7, 2020
MATH-VisionOriginalFebruary 22, 2024
MathVerseMINIMarch 21, 2024
MathVistaMINIOctober 3, 2023
MCP AtlasPublic, April 2026 updateApril 8, 2026
MedXpertQAMMJanuary 30, 2025