OWID: Share of FrontierMath problems solved correctly by AI models
The FrontierMath benchmark tests AI models on research-level mathematics problems that take expert mathematicians hours or days to solve. Source: Epoch AI (2026), compiled by Our World in Data. 104 values; read with owid_series(slug='ai-frontiermath-over-time').
Coverage
Observation dates are unavailable in the catalogue metadata.
104 rows reported at indexing.
Dimensions and measures
Open the source MCP to inspect available dimensions, measures, and filter values.