OpenAI · model
GPT-5 benchmark scores
As of 2026-11-18, GPT-5 (OpenAI) has measured scores on 24 benchmarks and estimated scores on 122 more.
| Benchmark | Score | Source |
|---|---|---|
| AA-GPQA Diamond | 85.4% | measured |
| AA-HLE | 28.5% | measured |
| AA-SciCode | 39.3% | estimated ± 3.4 pp, high confidence |
| Artificial Analysis Intelligence Index | 23.0% | measured |
| CritPt | 5.7% | measured |
| GDPval-AA | 20.9% | measured |
| GPQA Diamond | 88.4% | estimated ± 1.8 pp, high confidence |
| HLE | 30.5% | estimated ± 8.5 pp, medium confidence |
| IFBench | 71.1% | estimated ± 6.6 pp, medium confidence |
| AA-LCR | 78.2% | measured |
| LiveCodeBench v6 | 89.1% | estimated ± 2.2 pp, medium confidence |
| AA-Omniscience Accuracy | 40.3% | measured |
| SciCode | 48.5% | estimated ± 2.9 pp, medium confidence |
| BrowseComp | 62.7% | estimated ± 6.3 pp, low confidence |
| HLE w/ tools | 43.9% | estimated ± 5.3 pp, low confidence |
| IFEval | 92.9% | estimated ± 2.8 pp, high confidence |
| LongBench v2 | 64.2% | estimated ± 5.5 pp, medium confidence |
| τ²-bench results | 84.8% | measured |
| AA Coding Index | 37.8% | measured |
| AA-MMMU-Pro | 74.2% | measured |
| APEX-Agents | 14.7% | estimated ± 9.4 pp, low confidence |
| APEX-Agents-AA | 12.3% | estimated ± 3.6 pp, low confidence |
| BioMysteryBench (human-difficult) | 38.6% | estimated ± 6.8 pp, low confidence |
| DeepSearchQA | 72.9% | estimated ± 8.8 pp, low confidence |
| SWE-bench Verified | 76.2% | estimated ± 4.4 pp, high confidence |
| BFCL v4 | 72.4% | estimated ± 2.7 pp, high confidence |
| CyberGym | 45.9% | estimated ± 6.0 pp, low confidence |
| JobBench | 8.5% | measured |
| SWE-bench Pro | 54.6% | estimated ± 2.9 pp, medium confidence |
| WideResearch | 72.2% | estimated ± 1.2 pp, low confidence |
| AA Agentic Index | 15.6% | estimated ± 3.1 pp, high confidence |
| MMLU-Pro | 85.7% | estimated ± 4.8 pp, high confidence |
| AA-IFBench | 73.1% | measured |
| FailureSensorIQ | 66.1% | estimated ± 2.1 pp, high confidence |
| GPQA | 87.4% | estimated ± 4.5 pp, high confidence |
| MMLU | 90.5% | estimated ± 0.9 pp, medium confidence |
| Vals GPQA Diamond | 85.6% | measured |
| Vals LiveCodeBench | 85.9% | measured |
| Vals MMLU-Pro | 86.5% | measured |
| MMMLU | 88.6% | estimated ± 0.7 pp, medium confidence |
| MMMU | 82.9% | estimated ± 2.7 pp, high confidence |
| Chess Puzzles | 36.0% | measured |
| AA MMLU-Pro | 87.1% | measured |
| Gert Labs | 50.2% | estimated ± 8.4 pp, low confidence |
| AA EnterpriseOps-Gym | 34.8% | estimated ± 1.2 pp, low confidence |
| AA Tau3 Banking | 26.7% | estimated ± 4.1 pp, medium confidence |
| ApprenticeBench | 0.0% | estimated ± 6.0 pp, low confidence |
| ARC-AGI-1 | 79.2% | estimated ± 6.5 pp, medium confidence |
| ARC-AGI-2 | 40.1% | estimated ± 13.8 pp, low confidence |
| CursorBench 3.1 | 42.8% | estimated ± 5.8 pp, low confidence |
| CursorBench 3.2 | 52.2% | estimated ± 5.4 pp, low confidence |
| EBR-bench | 12.7% | measured |
| FrontierCode 1.1 Main | 12.9% | estimated ± 4.6 pp, medium confidence |
| FrontierSWE v2 | 6.1% | estimated ± 9.5 pp, low confidence |
| Furniture Assembly | 28.0% | estimated ± 6.7 pp, medium confidence |
| MaxxBench | 39.3% | estimated ± 4.3 pp, low confidence |
| MazeBench | 0.0% | estimated ± 5.1 pp, low confidence |
| MirrorCode | 4.4% | measured |
| Mystery Game Puzzles | 23.0% | measured |
| NYT Connections | 79.5% | estimated ± 12.6 pp, low confidence |
| OfficeQA Pro | 50.3% | estimated ± 6.8 pp, low confidence |
| OSWorld-Verified | 66.5% | estimated ± 8.4 pp, low confidence |
| ParseBench-1 | 58.2% | estimated ± 6.2 pp, medium confidence |
| Vals SWE-bench | 69.0% | measured |
| VulcanBench v3 | 84.2% | estimated ± 5.6 pp, low confidence |
| CWE-Bench-AA | 28.2% | estimated ± 10.1 pp, low confidence |
| DeepsecBench-AA | 0.4% | estimated ± 8.9 pp, low confidence |
| HealthBench Professional (length-adjusted, Artificial Analysis) | 52.8% | estimated ± 4.3 pp, low confidence |
| HealthBench Professional (raw, Artificial Analysis) | 44.5% | estimated ± 0.8 pp, low confidence |
| AA ITBench | 36.3% | estimated ± 5.1 pp, low confidence |
| Chartography (tools) | 79.7% | estimated ± 4.1 pp, low confidence |
| CursorBench 4.0 | 30.8% | estimated ± 3.4 pp, medium confidence |
| CWE-bench v1 | 29.9% | estimated ± 6.4 pp, low confidence |
| DeepSWE | 56.3% | estimated ± 7.2 pp, low confidence |
| DRACO | 68.8% | estimated ± 2.0 pp, low confidence |
| Epoch Automation Reports | 11.1% | estimated ± 8.4 pp, low confidence |
| FrontierPhysics | 2.6% | estimated ± 4.2 pp, high confidence |
| HLE w/o tools | 35.7% | estimated ± 4.9 pp, high confidence |
| PostTrainBench v1.1 | 14.5% | estimated ± 3.7 pp, medium confidence |
| SpireClimb | 0.0% | estimated ± 10.3 pp, low confidence |
| StealthMark | 35.8% | estimated ± 8.7 pp, low confidence |
| Toolathlon-Verified | 46.0% | estimated ± 3.1 pp, medium confidence |
| GDP.pdf | 16.4% | estimated ± 4.1 pp, low confidence |
| ARC-AGI-3 | 0.0% | estimated ± 8.6 pp, low confidence |
| ARC-AGI-3 Provider Adapter | 2.7% | estimated ± 7.3 pp, low confidence |
| SkillsBench official leaderboard (BenchFlow) | 55.3% | estimated ± 6.2 pp, medium confidence |
| CompileBench | 86.7% | measured |
| OTelBench | 25.6% | estimated ± 4.1 pp, medium confidence |
| Vibe Code Bench | 20.1% | measured |
| FACTS Multimodal | 42.1% | estimated ± 5.6 pp, medium confidence |
| Hermes Bench | 51.6% | estimated ± 2.9 pp, medium confidence |
| Hermes Index | 19.5% | estimated ± 8.7 pp, low confidence |
| ParseBench-2 | 63.4% | estimated ± 6.2 pp, medium confidence |
| CharXiv | 80.4% | estimated ± 6.3 pp, low confidence |
| CharXiv w/o tools | 79.7% | estimated ± 2.5 pp, high confidence |
| ExploitBench | 0.0% | estimated ± 6.4 pp, low confidence |
| ExploitGym | 0.3% | estimated ± 10.6 pp, low confidence |
| AI2D_TEST | 90.4% | estimated ± 2.0 pp, medium confidence |
| C-Eval | 92.4% | estimated ± 1.0 pp, medium confidence |
| Claw-Eval | 50.4% | estimated ± 4.5 pp, low confidence |
| CountBench | 94.4% | estimated ± 2.9 pp, medium confidence |
| DynaMath | 85.0% | estimated ± 3.0 pp, medium confidence |
| ERQA | 60.2% | estimated ± 3.1 pp, high confidence |
| MathVision | 84.2% | estimated ± 3.1 pp, high confidence |
| MCP Atlas | 22.6% | estimated ± 8.9 pp, low confidence |
| MMLU-Redux | 96.0% | estimated ± 0.8 pp, medium confidence |
| MMMU-Pro | 75.5% | estimated ± 1.5 pp, high confidence |
| NL2Repo | 39.7% | estimated ± 4.7 pp, high confidence |
| PostTrainBench v1 | 19.0% | estimated ± 3.1 pp, medium confidence |
| QwenClawBench | 55.9% | estimated ± 6.0 pp, low confidence |
| RealWorldQA | 82.5% | estimated ± 2.9 pp, medium confidence |
| ScreenSpot Pro | 74.0% | estimated ± 10.8 pp, low confidence |
| SuperGPQA | 81.8% | estimated ± 9.3 pp, medium confidence |
| Toolathlon | 30.0% | estimated ± 6.3 pp, medium confidence |
| V* | 100.0% | estimated ± 7.5 pp, medium confidence |
| Video-MME (w/o subtitle) | 82.5% | estimated ± 1.7 pp, medium confidence |
| Video-MME (with subtitle) | 84.6% | estimated ± 3.7 pp, medium confidence |
| AA LiveCodeBench | 84.6% | measured |
| HealthBench Hard | 24.8% | estimated ± 8.6 pp, low confidence |
| MedXpertQA (MM) | 70.2% | estimated ± 6.5 pp, low confidence |
| React Native Evals | 68.6% | estimated ± 3.0 pp, low confidence |
| ResearchClawBench | 15.1% | estimated ± 2.2 pp, medium confidence |
| SWE-Rebench | 32.3% | estimated ± 6.5 pp, low confidence |
| PostTrainBench | 12.1% | estimated ± 3.7 pp, medium confidence |
| DAYJOB | 1.3% | estimated ± 3.2 pp, high confidence |
| DAYJOB Healthcare | 0.7% | estimated ± 4.2 pp, high confidence |
| BioMysteryBench (human-solvable) | 81.8% | estimated ± 0.6 pp, low confidence |
| HealthBench (raw) | 37.3% | estimated ± 6.0 pp, low confidence |
| HealthBench (length-adjusted) | 49.2% | estimated ± 3.4 pp, low confidence |
| HealthBench Professional | 56.7% | estimated ± 4.2 pp, low confidence |
| HealthBench Professional (raw) | 48.6% | estimated ± 4.5 pp, medium confidence |
| HLE-Verified | 1.2% | estimated ± 4.6 pp, low confidence |
| LABBench2 | 81.2% | estimated ± 1.8 pp, low confidence |
| ScarfBench | 24.4% | estimated ± 9.1 pp, low confidence |
| BlenderWorld (text-only) | 70.2% | estimated ± 2.8 pp, medium confidence |
| Cookie-Bench | 77.4% | estimated ± 1.3 pp, medium confidence |
| STEMBench | 2.5% | estimated ± 0.2 pp, medium confidence |
| SimpleQA | 46.0% | estimated ± 4.3 pp, low confidence |
| BabyVision w/ Python | 87.1% | estimated ± 4.0 pp, low confidence |
| OpenHarmony Bench v1.0 | 42.0% | estimated ± 2.8 pp, medium confidence |
| SEC-Bench Pro | 26.9% | estimated ± 7.5 pp, low confidence |
| MRCRv2 | 85.5% | estimated ± 11.3 pp, low confidence |
| MMLongBench-Doc | 61.3% | estimated ± 2.6 pp, low confidence |
| BioBasicsBench | 33.9% | estimated ± 2.8 pp, low confidence |
| MMMU-Pro w/ Python | 77.5% | estimated ± 2.0 pp, high confidence |
| ExploitBench Internal Port | 0.2% | estimated ± 1.4 pp, medium confidence |