GPT-4o (May '24) — benchmark results

Provider: OpenAI. Released 2024-05-13. Access: API.

Unified ELO 1492 ± 41, rank #872 of 1841 rated models, from 7 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AA MATH-50079.13Accuracy (%)45.6
AA MMLU-Pro73.95Accuracy (%)44.5
Epoch AI - Scicode30.9Score43.9
AA LiveCodeBench33.44Pass@1 (%)39.9
AA GPQA Diamond52.63Accuracy (%)29.5
Artificial Analysis Intelligence Index8.6Intelligence Index28.1
AA Humanity's Last Exam2.76Accuracy (%)0.5

Interactive version: theaggregate.ai/model?slug=gpt-4o-may-24 · How It Works · Data refreshed daily, snapshot 2026-07-25.