OpenAI · model
o4-mini (high) benchmark scores
As of 2026-10-07, o4-mini (high) (OpenAI) has measured scores on 2 benchmarks and estimated scores on 4 more.
| Benchmark | Score | Source |
|---|---|---|
| AIME26 | 95.5% | estimated ± 0.6 pp, medium confidence |
| FrontierMath v2 (Tier 4) | 6.3% | measured |
| FrontierMath v2 (Tiers 1-3) | 24.8% | measured |
| HMMT Nov 2025 | 93.2% | estimated ± 2.6 pp, medium confidence |
| MMAnswerBench | 83.3% | estimated ± 1.3 pp, medium confidence |
| FrontierMath (legacy) | 25.6% | estimated ± 0.7 pp, low confidence |