GPT-5.4 Pro (2026-03-05): benchmark results
Provider: OpenAI. Released 2026-03-05. Access: API.
Aggregate Rating 1820 ± 12, rank #61 of 10916 rated models, from 16 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Epoch AI - GPQA Diamond | 94.6 | Accuracy (%) | 98.2 |
| SEAL - VISTA | 53.89 | Score | 96.8 |
| SEAL - TutorBench | 56.62 | Score | 96.2 |
| Epoch AI - Critpt | 30 | Score | 93.8 |
| SEAL - Humanity's Last Exam (Text Only) | 45.32 | Score | 93.8 |
| Chess Puzzles (Epoch AI) | 59 | Accuracy (%) | 93.5 |
| SEAL - MultiNRC | 62.27 | Score | 93.2 |
| EnigmaEval | 23.82 | Score (imported) | 92.2 |
| SEAL - MASK | 91.73 | Score | 90.9 |
| SEAL - Humanity's Last Exam | 44.32 | Score | 90.7 |
| FrontierMath - Tiers 1-3 (v2) | 82.46 | Accuracy (%, 285 private v2 problems) | 90.6 |
| SEAL - MultiChallenge | 69.23 | Score | 77.1 |
Interactive version: theaggregate.ai/model?slug=gpt-5-4-pro-2026-03-05 · How It Works · Data refreshed daily, snapshot 2026-11-08.