OpenAI · model
GPT-4.1 benchmark scores
As of 2026-10-07, GPT-4.1 (OpenAI) has measured scores on 17 benchmarks and estimated scores on 72 more.
| Benchmark | Score | Source |
|---|---|---|
| AA-GPQA Diamond | 66.6% | measured |
| AA-HLE | 4.2% | measured |
| AA-SciCode | 35.6% | estimated ± 3.9 pp, high confidence |
| AIME26 | 95.5% | estimated ± 0.6 pp, low confidence |
| Artificial Analysis Intelligence Index | 12.7% | measured |
| CritPt | 0.0% | measured |
| GDPval-AA | 9.3% | estimated ± 8.8 pp, low confidence |
| GPQA Diamond | 66.3% | estimated ± 0.4 pp, high confidence |
| HLE | 10.3% | estimated ± 8.2 pp, low confidence |
| IFBench | 56.2% | estimated ± 6.9 pp, medium confidence |
| AA-LCR | 68.3% | measured |
| LiveCodeBench v6 | 76.7% | estimated ± 8.1 pp, low confidence |
| AA-Omniscience Accuracy | 27.8% | measured |
| SciCode | 35.9% | estimated ± 3.8 pp, high confidence |
| BrowseComp | 58.6% | estimated ± 5.7 pp, low confidence |
| IFEval | 87.4% | measured |
| LongBench v2 | 61.5% | estimated ± 5.9 pp, medium confidence |
| τ²-bench results | 47.1% | measured |
| AA Coding Index | 25.5% | estimated ± 5.1 pp, medium confidence |
| AA-MMMU-Pro | 61.2% | measured |
| APEX-Agents-AA | 1.6% | estimated ± 11.4 pp, low confidence |
| BioMysteryBench (human-difficult) | 33.4% | estimated ± 5.9 pp, low confidence |
| SWE-bench Verified | 54.6% | measured |
| BFCL v4 | 45.5% | estimated ± 9.8 pp, low confidence |
| CyberGym | 15.6% | estimated ± 10.0 pp, low confidence |
| SWE-bench Pro | 28.8% | estimated ± 3.2 pp, high confidence |
| τ³-bench results | 100.0% | estimated ± 8.9 pp, low confidence |
| WideResearch | 12.3% | estimated ± 5.9 pp, low confidence |
| AA Agentic Index | 5.7% | estimated ± 6.4 pp, low confidence |
| MMLU-Pro | 76.8% | estimated ± 3.0 pp, high confidence |
| FrontierMath v2 (Tier 4) | 0.0% | measured |
| FrontierMath v2 (Tiers 1-3) | 5.5% | measured |
| GPQA | 66.3% | measured |
| AA-IFBench | 43.0% | measured |
| Gert Labs | 25.7% | measured |
| AA EnterpriseOps-Gym | 2.2% | estimated ± 3.6 pp, low confidence |
| ApprenticeBench | 0.0% | estimated ± 3.1 pp, low confidence |
| ARC-AGI-1 | 8.3% | estimated ± 5.4 pp, low confidence |
| ARC-AGI-2 | 0.0% | estimated ± 11.6 pp, low confidence |
| FrontierCode 1.1 Main | 0.1% | estimated ± 2.7 pp, low confidence |
| OfficeQA Pro | 50.3% | estimated ± 7.2 pp, low confidence |
| OSWorld-Verified | 51.8% | estimated ± 5.9 pp, low confidence |
| Vals GPQA Diamond | 50.8% | estimated ± 4.8 pp, medium confidence |
| Vals LiveCodeBench | 32.0% | estimated ± 9.9 pp, low confidence |
| Vals MMLU-Pro | 12.4% | estimated ± 1.2 pp, medium confidence |
| Vals SWE-bench | 41.4% | estimated ± 6.8 pp, low confidence |
| AA ITBench | 95.6% | estimated ± 3.6 pp, low confidence |
| HLE w/o tools | 6.5% | estimated ± 4.9 pp, medium confidence |
| OSWorld 2.0 | 0.0% | estimated ± 4.5 pp, low confidence |
| SWE Multilingual | 56.7% | estimated ± 2.6 pp, low confidence |
| Vibe Code Bench | 0.6% | estimated ± 8.1 pp, low confidence |
| CharXiv | 70.4% | estimated ± 7.5 pp, low confidence |
| CharXiv w/o tools | 70.1% | estimated ± 2.4 pp, medium confidence |
| C-Eval | 78.2% | estimated ± 0.9 pp, low confidence |
| Claw-Eval | 45.8% | estimated ± 6.8 pp, low confidence |
| HMMT Nov 2025 | 99.9% | estimated ± 2.6 pp, low confidence |
| MathVision | 81.5% | estimated ± 4.9 pp, high confidence |
| MCP Atlas | 43.5% | estimated ± 14.8 pp, low confidence |
| MCP-Tasks | 96.8% | estimated ± 8.2 pp, low confidence |
| MMAnswerBench | 81.9% | estimated ± 1.3 pp, low confidence |
| MMLU-ProX | 72.7% | measured |
| MMLU-Redux | 95.1% | estimated ± 1.1 pp, medium confidence |
| MMMU-Pro | 70.4% | estimated ± 4.4 pp, medium confidence |
| NL2Repo | 8.6% | estimated ± 6.1 pp, low confidence |
| NOVA-63 | 51.5% | estimated ± 1.3 pp, low confidence |
| ScreenSpot Pro | 62.3% | estimated ± 10.7 pp, low confidence |
| SuperGPQA | 72.7% | estimated ± 7.5 pp, medium confidence |
| Toolathlon | 23.9% | estimated ± 7.2 pp, low confidence |
| ERQA | 0.7% | estimated ± 3.9 pp, medium confidence |
| HealthBench Hard | 0.0% | estimated ± 5.1 pp, low confidence |
| MedXpertQA (MM) | 5.0% | estimated ± 6.5 pp, low confidence |
| MMLU-Pro (Arcee) | 77.1% | estimated ± 4.0 pp, low confidence |
| ResearchClawBench | 12.0% | estimated ± 2.2 pp, low confidence |
| SWE-Rebench | 57.0% | estimated ± 3.6 pp, low confidence |
| ARC-AGI-3 | 0.0% | estimated ± 2.1 pp, medium confidence |
| FrontierMath (legacy) | 6.6% | estimated ± 0.7 pp, low confidence |
| BioMysteryBench (human-solvable) | 81.7% | estimated ± 0.6 pp, low confidence |
| HealthBench (raw) | 6.6% | estimated ± 6.4 pp, low confidence |
| HealthBench (length-adjusted) | 53.9% | estimated ± 3.7 pp, low confidence |
| HealthBench Professional | 12.4% | estimated ± 4.8 pp, low confidence |
| HealthBench Professional (raw) | 20.9% | estimated ± 5.0 pp, low confidence |
| HLE-Verified | 15.3% | estimated ± 2.0 pp, low confidence |
| LABBench2 | 47.0% | estimated ± 1.8 pp, low confidence |
| MMMU | 73.9% | estimated ± 0.9 pp, high confidence |
| LiveCodeBench | 68.4% | estimated ± 10.2 pp, low confidence |
| MMLU | 90.2% | measured |
| MMMU-Pro w/ Python | 69.2% | estimated ± 1.6 pp, medium confidence |
| RealWorldQA | 76.5% | estimated ± 1.2 pp, low confidence |
| OmniDocBench 1.5 | 0.0% | estimated ± 6.4 pp, low confidence |