Acropolis

GPT-5.2 Pro

OpenAIGrade C. Evidence grade C of A to D: thin.ProprietaryReleased 11 Dec 2025
Compare with peers

Benchmark appearances

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

BenchmarkRankScoreCostStatusObserved
ARC-AGI-2

Epoch AI Benchmarking Hub · high

#118 / 299
54.2%$15.72StandardNot stated
ARC-AGI-2

Epoch AI Benchmarking Hub · medium

#139 / 299
38.5%$8.99StandardNot stated

Reasoning effort scaling

Same benchmark and harness measured across multiple reasoning effort settings.

  • ARC-AGI-2

    Epoch AI Benchmarking Hub

    Monotone. Evidence grade A of A to D: strongest.
    medium

    38.5

    $8.99

    high

    54.2

    $15.72