O3 (Medium) — benchmark results
O3 evaluated at the medium reasoning-effort setting. Provider: OpenAI. Released 2025-04-16. Access: API.
Unified ELO 1791 ± 32, rank #149 of 1776 rated models, from 35 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Gapminder AI Worldview | 93.9 | Correct Rate (%) | 100 |
| HAL AssistantBench | 38.81 | Accuracy (%) | 100 |
| HAL ScienceAgentBench | 33.33 | Accuracy (%) | 100 |
| Step Game (Lechmazur) | 5.32 | TrueSkill μ | 98.6 |
| HAL SciCode | 9.23 | Accuracy (%) | 96.7 |
| SEAL - EnigmaEval | 13.09 | Score | 91.9 |
| IneqMath | 37 | Overall Accuracy (self-reported) | 90.2 |
| NYT Connections Older Models | 73 | Score (%) | 89.8 |
| LLM Chess (Saplin) | 777.6 | ELO | 89.3 |
| LisanBench | 0.23 | Mean Path Length / Current Maximum | 87.9 |
| EnigmaEval | 13.09 | Score (self-reported) | 86.8 |
| HAL TAU-bench Airline | 54 | Accuracy (%) | 82.1 |
Interactive version: theaggregate.ai/model?slug=o3-medium · How the rankings work · Data refreshed daily, snapshot 2026-07-22.