Acropolis

GPT-5 Pro

OpenAIGrade C. Evidence grade C of A to D: thin.ProprietaryReleased 7 Oct 2025
Compare with peers

Benchmark appearances

Best verified run per board, with confidence intervals and Pareto frontier status.

BenchmarkRankScoreCostStatusObserved
HLE

Epoch AI Benchmarking Hub · Default

#12 / 56
31.6%[28.1% to 35.2%]UnknownStandardNot stated
ARC-AGI-2

Epoch AI Benchmarking Hub · Default

#168 / 299
18.3%$7.14StandardNot stated