OpenAI · model
GPT-5.4 nano benchmark scores
As of 2026-10-07, GPT-5.4 nano (OpenAI) has measured scores on 33 benchmarks and estimated scores on 90 more.
| Benchmark | Score | Source |
|---|---|---|
| AA-GPQA Diamond | 81.7% | measured |
| AA-HLE | 28.3% | measured |
| AA-SciCode | 47.2% | measured |
| AIME26 | 95.5% | estimated ± 0.6 pp, medium confidence |
| Artificial Analysis Intelligence Index | 20.7% | measured |
| CritPt | 9.3% | measured |
| GDPval-AA | 22.5% | measured |
| GPQA Diamond | 82.8% | estimated ± 0.1 pp, high confidence |
| HLE | 37.7% | measured |
| IFBench | 74.5% | estimated ± 6.9 pp, medium confidence |
| AA-LCR | 76.7% | measured |
| LiveCodeBench v6 | 87.7% | estimated ± 3.1 pp, high confidence |
| AA-Omniscience Accuracy | 25.7% | measured |
| SciCode | 44.7% | estimated ± 2.9 pp, medium confidence |
| Terminal-Bench 2.1 | 38.8% | estimated ± 6.9 pp, low confidence |
| BrowseComp | 62.8% | estimated ± 3.1 pp, medium confidence |
| HLE w/ tools | 43.1% | estimated ± 6.1 pp, low confidence |
| IFEval | 93.7% | estimated ± 1.3 pp, high confidence |
| LongBench v2 | 61.5% | estimated ± 5.9 pp, medium confidence |
| VITA-Bench | 37.7% | estimated ± 9.9 pp, low confidence |
| τ²-bench results | 92.5% | measured |
| AA Coding Index | 56.1% | measured |
| AA-MMMU-Pro | 65.4% | measured |
| APEX-Agents | 14.7% | estimated ± 0.6 pp, low confidence |
| APEX-Agents-AA | 24.9% | measured |
| BioMysteryBench (human-difficult) | 31.9% | estimated ± 5.9 pp, low confidence |
| DeepSearchQA | 82.6% | estimated ± 4.5 pp, high confidence |
| SWE-bench Verified | 76.6% | estimated ± 4.4 pp, high confidence |
| AutomationBench | 28.4% | estimated ± 9.7 pp, low confidence |
| BFCL v4 | 73.5% | estimated ± 2.6 pp, high confidence |
| CyberGym | 59.2% | estimated ± 5.3 pp, low confidence |
| JobBench | 29.5% | estimated ± 6.8 pp, medium confidence |
| SWE-bench Pro | 56.6% | estimated ± 2.1 pp, high confidence |
| τ³-bench results | 69.5% | estimated ± 1.8 pp, medium confidence |
| WideResearch | 71.3% | estimated ± 5.6 pp, low confidence |
| AA Agentic Index | 17.7% | measured |
| MMLU-Pro | 84.8% | estimated ± 2.5 pp, medium confidence |
| FrontierMath v2 (Tier 4) | 6.3% | measured |
| FrontierMath v2 (Tiers 1-3) | 25.9% | measured |
| GPQA | 82.8% | measured |
| AA-IFBench | 75.9% | measured |
| Gert Labs | 47.2% | estimated ± 7.2 pp, medium confidence |
| AA-AnalystAgent | 15.5% | estimated ± 12.0 pp, low confidence |
| AA EnterpriseOps-Gym | 42.2% | estimated ± 3.6 pp, medium confidence |
| AA Harvey LAB | 80.2% | estimated ± 1.2 pp, low confidence |
| ApprenticeBench | 0.0% | estimated ± 5.0 pp, low confidence |
| ARC-AGI-1 | 51.5% | measured |
| ARC-AGI-2 | 5.7% | measured |
| CursorBench 3.1 | 37.1% | estimated ± 3.4 pp, low confidence |
| CursorBench 3.2 | 41.0% | estimated ± 4.4 pp, medium confidence |
| Terminal-Bench 3.0 | 0.7% | estimated ± 6.9 pp, low confidence |
| FrontierCode 1.1 Main | 25.8% | estimated ± 3.8 pp, medium confidence |
| FrontierSWE v2 | 6.3% | estimated ± 9.5 pp, low confidence |
| OfficeQA Pro | 50.3% | estimated ± 7.2 pp, low confidence |
| OSWorld-Verified | 39.0% | measured |
| Vals GPQA Diamond | 77.5% | measured |
| Vals LiveCodeBench | 84.0% | measured |
| Vals MMLU-Pro | 77.2% | measured |
| Vals SWE-bench | 69.8% | measured |
| Terminal-Bench 2.1 (Vals AI) | 41.6% | measured |
| VulcanBench v3 | 83.0% | estimated ± 5.6 pp, low confidence |
| AA AutomationBench | 3.9% | estimated ± 7.0 pp, medium confidence |
| GDP.pdf | 9.4% | estimated ± 5.4 pp, medium confidence |
| AA ITBench | 27.9% | estimated ± 3.6 pp, low confidence |
| AA Tau3 Banking | 22.0% | estimated ± 5.8 pp, medium confidence |
| AA Terminal-Bench 2.1 | 34.0% | estimated ± 4.4 pp, medium confidence |
| AA Terminal-Bench 4.0 | 0.0% | estimated ± 14.0 pp, low confidence |
| CursorBench 4.0 | 19.8% | estimated ± 1.6 pp, medium confidence |
| CWE-bench v1 | 32.2% | estimated ± 6.9 pp, low confidence |
| DeepSWE | 24.7% | estimated ± 6.4 pp, low confidence |
| HLE w/o tools | 24.3% | measured |
| OSWorld 2.0 | 0.0% | estimated ± 3.6 pp, medium confidence |
| PostTrainBench v1.1 | 19.0% | estimated ± 2.7 pp, high confidence |
| Toolathlon-Verified | 44.8% | estimated ± 2.3 pp, low confidence |
| Vibe Code Bench | 26.1% | measured |
| CharXiv | 61.0% | estimated ± 5.4 pp, medium confidence |
| CharXiv w/o tools | 69.4% | estimated ± 1.8 pp, low confidence |
| ExploitGym | 0.0% | estimated ± 12.2 pp, low confidence |
| C-Eval | 93.2% | estimated ± 0.8 pp, low confidence |
| Claw-Eval | 63.4% | estimated ± 5.7 pp, medium confidence |
| DeepPlanning | 37.9% | estimated ± 13.1 pp, low confidence |
| HMMT Nov 2025 | 93.2% | estimated ± 2.6 pp, medium confidence |
| MathVision | 66.5% | estimated ± 3.6 pp, medium confidence |
| MCP Atlas | 56.1% | measured |
| MCP-Tasks | 76.0% | estimated ± 0.6 pp, low confidence |
| MMAnswerBench | 83.3% | estimated ± 1.3 pp, medium confidence |
| MMLU-Redux | 94.7% | estimated ± 1.1 pp, medium confidence |
| MMMU-Pro | 66.1% | measured |
| NL2Repo | 40.7% | estimated ± 4.7 pp, high confidence |
| QwenClawBench | 54.9% | estimated ± 2.6 pp, high confidence |
| ScreenSpot Pro | 46.7% | estimated ± 10.0 pp, low confidence |
| SuperGPQA | 70.7% | estimated ± 7.5 pp, medium confidence |
| Terminal-Bench 2.0 | 46.3% | measured |
| Toolathlon | 35.5% | measured |
| VideoMMMU | 84.1% | estimated ± 1.0 pp, medium confidence |
| ERQA | 5.3% | estimated ± 3.9 pp, medium confidence |
| HealthBench Hard | 0.0% | estimated ± 5.1 pp, low confidence |
| MedXpertQA (MM) | 41.2% | estimated ± 2.4 pp, medium confidence |
| MedXpertQA (Text) | 39.3% | estimated ± 8.9 pp, low confidence |
| MMLU-Pro (Arcee) | 100.0% | estimated ± 4.0 pp, low confidence |
| React Native Evals | 65.4% | estimated ± 2.4 pp, low confidence |
| ResearchClawBench | 16.8% | estimated ± 1.8 pp, medium confidence |
| SWE-Rebench | 52.2% | estimated ± 6.5 pp, low confidence |
| ARC-AGI-3 | 0.0% | estimated ± 2.1 pp, medium confidence |
| FrontierMath (legacy) | 26.7% | estimated ± 0.7 pp, low confidence |
| BioMysteryBench (human-solvable) | 82.8% | estimated ± 0.6 pp, low confidence |
| DRACO | 70.8% | estimated ± 3.0 pp, medium confidence |
| FrontierCode 1.1 Extended | 15.6% | estimated ± 2.4 pp, low confidence |
| HealthBench (raw) | 37.4% | estimated ± 6.4 pp, low confidence |
| HealthBench (length-adjusted) | 37.4% | estimated ± 2.5 pp, low confidence |
| HealthBench Professional | 49.8% | estimated ± 1.3 pp, low confidence |
| HealthBench Professional (raw) | 48.6% | estimated ± 5.0 pp, low confidence |
| HLE-Verified | 35.3% | estimated ± 0.7 pp, low confidence |
| LABBench2 | 11.5% | estimated ± 1.3 pp, low confidence |
| MMMU | 76.1% | estimated ± 0.9 pp, high confidence |
| LiveCodeBench | 100.0% | estimated ± 8.9 pp, low confidence |
| OpenHarmony Bench | 51.0% | estimated ± 2.4 pp, high confidence |
| SimpleVQA | 59.8% | estimated ± 8.2 pp, medium confidence |
| MMLU | 89.2% | estimated ± 1.1 pp, medium confidence |
| MMMU-Pro w/ Python | 69.5% | measured |
| RealWorldQA | 79.3% | estimated ± 1.2 pp, low confidence |
| Video-MME (with subtitle) | 83.4% | estimated ± 1.8 pp, low confidence |
| OmniDocBench 1.5 | 0.0% | estimated ± 5.9 pp, low confidence |