O4 Mini (2025-04-16): benchmark results
April 16, 2025 O4 Mini snapshot, tracked when sources report the dated API model. Provider: OpenAI. Released 2025-04-16. Access: API.
Aggregate Rating 1707 ± 8, rank #179 of 1656 rated models, from 1080 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Arcadia MMMU Multiple Choice | 79.46 | Accuracy (%) | 100 |
| Defects4J | 53.8 | Defects4J Plausible @1 (imported) | 100 |
| HELM MedHELM - Clinical Decision Support | 78.47 | Mean win rate | 100 |
| HELM MedHELM - Mental Health | 4.98 | Jury Score | 100 |
| HELM VHELM - Blink Perception - Visual Correspondence | 95.93 | Prefix Quasi-Exact Match (%) | 100 |
| HELM VHELM - Crossmodal 3600 | 3.96 | Prometheus Vision rating | 100 |
| HELM VHELM - Crossmodal 3600 - Chinese Locale | 4.05 | Prometheus Vision Rating (1-5) | 100 |
| HELM VHELM - Crossmodal 3600 - Cusco Quechua Locale | 4.2 | Prometheus Vision Rating (1-5) | 100 |
| HELM VHELM - Crossmodal 3600 - English Locale | 3.75 | Prometheus Vision Rating (1-5) | 100 |
| HELM VHELM - Crossmodal 3600 - Hindi Locale | 3.76 | Prometheus Vision Rating (1-5) | 100 |
| HELM VHELM - Crossmodal 3600 - Maori Locale | 4.09 | Prometheus Vision Rating (1-5) | 100 |
| HELM VHELM - Crossmodal 3600 - Spanish Locale | 3.97 | Prometheus Vision Rating (1-5) | 100 |
Interactive version: theaggregate.ai/model?slug=o4-mini-2025-04-16 · How It Works · Data refreshed daily, snapshot 2026-10-20.