O3 Mini (2025-01-31): benchmark results
January 31, 2025 O3 Mini snapshot, tracked when sources report the dated API model. Provider: OpenAI. Released 2025-01-31. Access: API.
Aggregate Rating 1683 ± 9, rank #240 of 1656 rated models, from 831 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| BIRD-CRITIC | 34.5 | Score | 100 |
| BigCodeBench-Hard | 33.1 | Instruct pass@1 (imported) | 100 |
| GT Bench (Graph Theory) - Easy | 98.66 | Accuracy (%; Easy subset: connectivity and bipartiteness alw | 100 |
| GT Bench (Graph Theory) - Hard | 75.09 | Accuracy (%; Hard subset: min-cost max-flow, cycle count, sp | 100 |
| HELM AIR-Bench 2024 - #39-40.31: Characterization of identity - Social behaviors | 100 | Refusal Rate (%) | 100 |
| HELM MedHELM - CLEAR | 83.06 | EM | 100 |
| HELM MedHELM - Medec | 68.68 | MedecFlagAcc | 100 |
| HELM MedHELM - Medical Research | 76.39 | Mean win rate | 100 |
| HELM MedHELM - SHC-GIP | 78.4 | EM | 100 |
| HoWToBench - Comment | 6.02 | Tree-of-Writing score: GPT-4o rates each content trait 1-10 | 100 |
| HoWToBench - Contract | 5.8 | Tree-of-Writing score: GPT-4o rates each content trait 1-10 | 100 |
| HoWToBench - Plan | 5.75 | Tree-of-Writing score: GPT-4o rates each content trait 1-10 | 100 |
Interactive version: theaggregate.ai/model?slug=o3-mini-2025-01-31 · How It Works · Data refreshed daily, snapshot 2026-10-20.