GPT-5.4 Nano (xHigh): benchmark results
GPT-5.4 Nano evaluated at the xhigh reasoning-effort setting. Provider: OpenAI. Released 2026-03-17. Access: API.
Aggregate Rating 1683 ± 13, rank #410 of 2168 rated models, from 124 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA IFBench | 75.92 | Accuracy (%) | 94.5 |
| AA Terminal-Bench Hard | 42.42 | Accuracy (%) | 89.2 |
| Korean CSAT 2026 (Easy Mode) - Mathematics | 100 | Points (out of 100) | 88.5 |
| CritPt | 9.3 | Accuracy (imported) | 82.4 |
| AA CritPt | 9.25 | Accuracy (%) | 78.2 |
| Korean CSAT 2026 (Easy Mode) - Korean | 98 | Points (out of 100) | 77.4 |
| LiveBench Table Join | 50.59 | Score | 75.7 |
| AA Long Context Reasoning | 76.67 | Accuracy (%) | 75.3 |
| AA Humanity's Last Exam | 28.27 | Accuracy (%) | 74.7 |
| Korean CSAT 2026 (Easy Mode) - Society and Culture | 43 | Points (out of 50) | 74 |
| Korean CSAT 2026 (Easy Mode) - Total | 424.5 | Points (out of 450) | 72.3 |
| Korean CSAT 2026 (Easy Mode) - Physics I | 39 | Points (out of 50) | 71.3 |
Interactive version: theaggregate.ai/model?slug=gpt-5-4-nano-xhigh · How It Works · Data refreshed daily, snapshot 2026-10-20.