OpenAI · model
o3-mini benchmark scores
As of 2026-11-21, o3-mini (OpenAI) has measured scores on 14 benchmarks and estimated scores on 63 more.
| Benchmark | Score | Source |
|---|---|---|
| AA-GPQA Diamond | 74.8% | measured |
| AA-HLE | 7.9% | measured |
| AA-SciCode | 32.9% | estimated ± 3.8 pp, high confidence |
| Artificial Analysis Intelligence Index | 12.5% | measured |
| CritPt | 0.3% | estimated ± 0.3 pp, high confidence |
| GDPval-AA | 0.1% | estimated ± 11.0 pp, low confidence |
| GPQA Diamond | 77.2% | estimated ± 0.1 pp, medium confidence |
| HLE | 14.1% | estimated ± 6.5 pp, medium confidence |
| IFBench | 72.6% | estimated ± 9.2 pp, medium confidence |
| LiveCodeBench v6 | 82.5% | estimated ± 2.2 pp, low confidence |
| AA-Omniscience Accuracy | 18.5% | estimated ± 2.4 pp, high confidence |
| SciCode | 32.3% | estimated ± 3.6 pp, high confidence |
| BrowseComp | 34.5% | estimated ± 13.0 pp, low confidence |
| IFEval | 93.9% | measured |
| τ²-bench results | 28.7% | measured |
| AA Coding Index | 24.4% | estimated ± 5.3 pp, medium confidence |
| APEX-Agents-AA | 1.6% | estimated ± 11.4 pp, low confidence |
| BioMysteryBench (human-difficult) | 13.5% | estimated ± 8.1 pp, low confidence |
| SWE-bench Verified | 49.3% | measured |
| BFCL v4 | 27.4% | estimated ± 8.5 pp, low confidence |
| SWE-bench Pro | 24.5% | estimated ± 4.5 pp, high confidence |
| AA Agentic Index | 1.3% | estimated ± 12.7 pp, low confidence |
| MMLU-Pro | 75.5% | estimated ± 4.0 pp, high confidence |
| AA-IFBench | 71.4% | estimated ± 8.8 pp, medium confidence |
| AIME 2024 | 87.3% | measured |
| BBH | 92.0% | estimated ± 2.4 pp, low confidence |
| FailureSensorIQ | 64.4% | estimated ± 2.1 pp, high confidence |
| GPQA | 77.2% | measured |
| MMLU | 86.9% | measured |
| Vals GPQA Diamond | 75.5% | measured |
| Vals LiveCodeBench | 71.5% | measured |
| Vals MMLU-Pro | 78.7% | measured |
| MATH-500 | 97.4% | estimated ± 1.3 pp, medium confidence |
| MMMLU | 85.9% | estimated ± 1.7 pp, high confidence |
| Chess Puzzles | 15.0% | measured |
| AIME 2025 | 81.3% | estimated ± 3.5 pp, high confidence |
| AA MMLU-Pro | 83.5% | estimated ± 1.1 pp, medium confidence |
| ARC-AGI-1 | 57.3% | estimated ± 6.5 pp, low confidence |
| ARC-AGI-2 | 0.0% | estimated ± 10.9 pp, low confidence |
| CursorBench 3.2 | 59.1% | estimated ± 7.2 pp, low confidence |
| EBR-bench | 0.0% | estimated ± 10.4 pp, low confidence |
| Furniture Assembly | 32.9% | estimated ± 9.3 pp, medium confidence |
| MaxxBench | 47.2% | estimated ± 8.8 pp, medium confidence |
| Mystery Game Puzzles | 7.0% | measured |
| NYT Connections | 53.7% | estimated ± 12.6 pp, low confidence |
| Vals SWE-bench | 34.9% | estimated ± 6.6 pp, low confidence |
| VulcanBench v3 | 78.2% | estimated ± 5.6 pp, low confidence |
| HealthBench Professional (length-adjusted, Artificial Analysis) | 49.6% | estimated ± 4.3 pp, low confidence |
| HealthBench Professional (raw, Artificial Analysis) | 46.2% | estimated ± 1.1 pp, low confidence |
| CursorBench 4.0 | 29.8% | estimated ± 4.5 pp, medium confidence |
| FrontierPhysics | 2.5% | estimated ± 5.0 pp, low confidence |
| HLE w/o tools | 24.6% | estimated ± 4.9 pp, high confidence |
| SpireClimb | 0.0% | estimated ± 9.3 pp, medium confidence |
| StealthMark | 32.1% | estimated ± 9.2 pp, medium confidence |
| ARC-AGI-3 | 1.3% | estimated ± 9.8 pp, low confidence |
| ARC-AGI-3 Provider Adapter | 2.7% | estimated ± 7.3 pp, low confidence |
| SkillsBench official leaderboard (BenchFlow) | 44.9% | estimated ± 6.2 pp, medium confidence |
| CompileBench | 55.6% | estimated ± 10.0 pp, low confidence |
| Vibe Code Bench | 0.0% | estimated ± 8.4 pp, low confidence |
| FACTS Multimodal | 34.4% | estimated ± 5.6 pp, low confidence |
| ExploitBench | 14.6% | estimated ± 7.2 pp, low confidence |
| C-Eval | 89.3% | estimated ± 1.0 pp, medium confidence |
| MMLU-Redux | 91.8% | estimated ± 1.2 pp, medium confidence |
| NL2Repo | 45.7% | estimated ± 7.9 pp, low confidence |
| SuperGPQA | 57.6% | estimated ± 8.5 pp, low confidence |
| React Native Evals | 65.8% | estimated ± 2.7 pp, low confidence |
| DAYJOB | 1.3% | estimated ± 3.2 pp, high confidence |
| DAYJOB Healthcare | 0.7% | estimated ± 4.2 pp, high confidence |
| HealthBench (length-adjusted) | 54.1% | estimated ± 3.7 pp, medium confidence |
| HealthBench Professional | 56.7% | estimated ± 4.2 pp, low confidence |
| BlenderWorld (text-only) | 70.0% | estimated ± 2.8 pp, low confidence |
| Cookie-Bench | 76.9% | estimated ± 1.3 pp, low confidence |
| SimpleQA | 26.0% | estimated ± 7.8 pp, medium confidence |
| LiveCodeBench | 57.6% | estimated ± 14.6 pp, low confidence |
| SEC-Bench Pro | 33.7% | estimated ± 9.3 pp, medium confidence |
| BioBasicsBench | 15.4% | estimated ± 10.2 pp, low confidence |
| ExploitBench Internal Port | 0.0% | estimated ± 1.4 pp, low confidence |