FrontierMath Benchmark

Explore score history ↗

FrontierMath is an AI benchmark in the Benchmark Radar catalog.

Released: 2024-08-11
Openness: unknown
Importer: Benchmark Saturation

Source provenance

Original evidence

FrontierMath results and reported scores

Reported benchmark scores · Benchmark Saturation

Reported scores: 5

Model	Organization	Reported value	Reported	Evidence
Gemini 3 Pro Preview	—	18.8	—	evidence
GPT-5.2 (high)	—	14.6	—	evidence
GPT-5.2 (medium)	—	14.6	—	evidence
GPT-5.2 (Pro)	—	29.2	—	evidence
GPT-5.2 (xhigh)	—	16.7	—	evidence

Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.

Which sources cite FrontierMath?

FrontierMath