Acropolis

GPT-5.1 Codex Max

OpenAIGrade B. Evidence grade B of A to D: good.ProprietaryReleased 19 Nov 2025
Compare with peers

Benchmark appearances

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

BenchmarkRankScoreCostStatusObserved
METR time horizon

METR time horizons · triframe_inspect/triframe_agent

#9 / 23Tied #1
223.7 min[134.3 min to 396.2 min]UnknownStandardNot stated
Terminal Bench

Epoch AI Benchmarking Hub · Codex CLI

#50 / 202
60.4%[55.2% to 65.7%]UnknownStandard19 Nov 2025