Acropolis

GPT-5.2 Codex

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

Benchmark appearances

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

BenchmarkRankScoreCostStatusObserved
Terminal Bench

Epoch AI Benchmarking Hub · Deep Agents

#30 / 202
66.5%[60.5% to 72.5%]UnknownStandard14 Jan 2026
LMArena WebDev

LMArena · Webdev react · Default

#124 / 140
1,329[1,318 to 1,340]UnknownStandard15 Nov 2026
LMArena WebDev

LMArena · Webdev HTML · Default

#128 / 155
1,347[1,332 to 1,363]UnknownStandard15 Nov 2026
LMArena WebDev

LMArena · Overall · Default

#129 / 156
1,339[1,330 to 1,348]UnknownStandard15 Nov 2026
LMArena WebDev

LMArena · Webdev · Default

#129 / 156
1,339[1,330 to 1,348]UnknownStandard15 Nov 2026