GPT-5.4
Benchmark scores
| Benchmark | Category | Rank | Score | Captured |
|---|---|---|---|---|
| AA Coding Index | code | #2 | 57.3% | 2026-05-02 |
| Terminal-Bench 2.0 | agents | #3 | 81.8% | 2026-03-07 |
| METR Time Horizon | agents | #5 | 5h42m | 2026-03-05 |
| Chatbot Arena | chat | #6 | 1469 | 2026-05-01 |
| OpenRouter · Weekly Usage | usage | #11 | #18 | 2026-05-02 |