GPT-5.4 Mini (2026-03-05): benchmark results
Provider: OpenAI. Released 2026-03-17. Access: API.
Aggregate Rating 1732 ± 35, rank #229 of 10916 rated models, from 20 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| OpenCompass Knowledge - Engineering (CompassBench 2604) | 96.2 | Score (%) | 95.5 |
| OpenCompass Language - NLP (CompassBench 2604) | 69.6 | Score (%) | 84.1 |
| OpenCompass Language - Instruction Following (CompassBench 2604) | 74.5 | Score (%) | 77.3 |
| OpenCompass Language - Creation (CompassBench 2604) | 72.1 | Score (%) | 68.2 |
| OpenCompass LLM - Knowledge (CompassBench 2604) | 91.5 | Score (%) | 65.9 |
| OpenCompass LLM - Language (CompassBench 2604) | 73 | Score (%) | 63.6 |
| OpenCompass Knowledge - Common Sense (CompassBench 2604) | 90 | Score (%) | 61.4 |
| OpenCompass Reasoning - Academic (CompassBench 2604) | 41.6 | Score (%) | 59.1 |
| OpenCompass Knowledge - Science (CompassBench 2604) | 90.8 | Score (%) | 52.3 |
| OpenCompass LLM - Reasoning (CompassBench 2604) | 55.9 | Score (%) | 50 |
| OpenCompass Knowledge - Social Science (CompassBench 2604) | 90.8 | Score (%) | 47.7 |
| OpenCompass Knowledge - Humanities (CompassBench 2604) | 89.6 | Score (%) | 36.4 |
Interactive version: theaggregate.ai/model?slug=gpt-5-4-mini-2026-03-05 · How It Works · Data refreshed daily, snapshot 2026-11-08.