Acropolis

GPT-4o (Nov '24)

OpenAIGrade C. Evidence grade C of A to D: thin.ProprietaryReleased 20 Nov 2024Consistent across benchmarks
Compare with peers

Domain profile

Median within-board percentile by capability domain across boards with sufficient peer overlap.

  • Agentic (1 boards): 0 (scale 0 to 100)Agentic (1 boards)0
  • Coding (1 boards): 0 (scale 0 to 100)Coding (1 boards)0
  • Finance (1 boards): 0 (scale 0 to 100)Finance (1 boards)0
  • General text (3 boards): 0 (scale 0 to 100)General text (3 boards)0
  • Math (2 boards): 3 (scale 0 to 100)Math (2 boards)3
  • Science (3 boards): 5 (scale 0 to 100)Science (3 boards)5

Benchmark appearances

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

BenchmarkRankScoreCostStatusObserved
GDPval

Epoch AI Benchmarking Hub · Default

#11 / 11
9.9%UnknownStandardNot stated
SWE-Bench verified

Epoch AI Benchmarking Hub · Default

#35 / 35
31%[26.9% to 35.1%]UnknownStandard11 Feb 2026
HLE

Epoch AI Benchmarking Hub · Default

#56 / 56
2.7%[1.5% to 4%]UnknownStandardNot stated
FrontierMath-2025-02-28-Private

Epoch AI Benchmarking Hub · Default

#96 / 101
0.3%[0% to 2.4%]UnknownStandard6 Mar 2025
GPQA diamond

Epoch AI Benchmarking Hub · Default

#264 / 338
47.9%[42.7% to 53%]UnknownStandard5 Feb 2025
ARC-AGI-2

Epoch AI Benchmarking Hub · Default

#281 / 299
0%$0.08StandardNot stated

Closest head-to-head rivals

Direct pairwise win rates for GPT-4o (Nov '24) across shared benchmark boards.