GPT-4o (Feb 2025): benchmark results
February 2025 GPT-4o snapshot kept separate from other GPT-4o rows because benchmark sources report distinct scores. Provider: OpenAI. Released 2024-05-13. Access: API.
Aggregate Rating 1596 ± 25, rank #1048 of 10916 rated models, from 6 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| PGG-Bench (Lechmazur) | 9.92 | TrueSkill μ | 44.4 |
| Generalization V1 (Lechmazur) | 2 | Avg Rank (lower is better) | 40 |
| NYT Connections Older Models | 10.3 | Score (%) | 35.7 |
| Confabulation Leaderboard (Lechmazur) | 26.73 | Confabulation rate % (lower is better) | 32.3 |
| Elimination Game (Lechmazur) | 3.51 | TrueSkill μ | 30.5 |
| Step Game (Lechmazur) | 1.05 | TrueSkill μ | 16.2 |
Interactive version: theaggregate.ai/model?slug=gpt-4o-feb-2025 · How It Works · Data refreshed daily, snapshot 2026-11-08.