OpenAI · model
GPT-5.6 Sol benchmark scores
As of 2026-10-07, GPT-5.6 Sol (OpenAI) has measured scores on 52 benchmarks and estimated scores on 76 more.
| Benchmark | Score | Source |
|---|---|---|
| AA-GPQA Diamond | 94.1% | measured |
| AA-HLE | 49.5% | measured |
| AA-SciCode | 57.1% | measured |
| AIME26 | 100.0% | estimated ± 0.6 pp, low confidence |
| Artificial Analysis Intelligence Index | 58.9% | measured |
| CritPt | 32.3% | measured |
| GDPval-AA | 55.6% | measured |
| GPQA Diamond | 94.6% | measured |
| HLE | 60.7% | estimated ± 5.5 pp, low confidence |
| IFBench | 71.2% | estimated ± 6.9 pp, medium confidence |
| AA-LCR | 84.0% | measured |
| LiveCodeBench v6 | 91.6% | estimated ± 2.1 pp, high confidence |
| AA-Omniscience Accuracy | 59.4% | measured |
| SciCode | 41.9% | estimated ± 2.9 pp, low confidence |
| Terminal-Bench 2.1 | 91.9% | measured |
| BrowseComp | 92.2% | measured |
| HLE w/ tools | 59.3% | estimated ± 4.6 pp, medium confidence |
| IFEval | 93.5% | estimated ± 1.3 pp, high confidence |
| LongBench v2 | 61.5% | estimated ± 5.9 pp, low confidence |
| VITA-Bench | 46.6% | estimated ± 9.9 pp, low confidence |
| τ²-bench results | 85.1% | measured |
| AA Coding Index | 77.4% | measured |
| AA-MMMU-Pro | 83.4% | measured |
| APEX-Agents | 44.7% | estimated ± 10.0 pp, low confidence |
| APEX-Agents-AA | 38.7% | estimated ± 2.9 pp, low confidence |
| BioMysteryBench (human-difficult) | 49.2% | estimated ± 5.9 pp, low confidence |
| DeepSearchQA | 93.4% | estimated ± 4.5 pp, high confidence |
| SWE-bench Verified | 85.7% | estimated ± 2.5 pp, high confidence |
| AutomationBench | 45.2% | estimated ± 5.8 pp, medium confidence |
| BFCL v4 | 77.9% | estimated ± 2.6 pp, medium confidence |
| CyberGym | 84.5% | measured |
| JobBench | 65.0% | estimated ± 4.9 pp, medium confidence |
| SWE-bench Pro | 64.6% | measured |
| τ³-bench results | 72.9% | estimated ± 2.1 pp, low confidence |
| WideResearch | 82.8% | estimated ± 4.6 pp, high confidence |
| AA Agentic Index | 50.5% | measured |
| MMLU-Pro | 87.4% | estimated ± 3.0 pp, medium confidence |
| FrontierMath v2 (Tier 4) | 83.0% | measured |
| FrontierMath v2 (Tiers 1-3) | 89.0% | measured |
| GPQA | 94.6% | measured |
| AA-IFBench | 72.7% | measured |
| Gert Labs | 75.5% | estimated ± 3.0 pp, low confidence |
| AA-AnalystAgent | 47.5% | measured |
| AA EnterpriseOps-Gym | 51.0% | estimated ± 1.8 pp, medium confidence |
| AA Harvey LAB | 93.5% | estimated ± 1.2 pp, medium confidence |
| ApprenticeBench | 26.0% | measured |
| ARC-AGI-1 | 96.5% | measured |
| ARC-AGI-2 | 92.5% | measured |
| CursorBench 3.1 | 61.5% | estimated ± 3.4 pp, medium confidence |
| CursorBench 3.2 | 67.2% | measured |
| Terminal-Bench 3.0 | 34.6% | measured |
| FrontierCode 1.1 Main | 50.5% | estimated ± 1.6 pp, medium confidence |
| FrontierSWE v2 | 32.2% | measured |
| OfficeQA Pro | 67.7% | estimated ± 7.2 pp, medium confidence |
| OSWorld-Verified | 82.2% | estimated ± 1.9 pp, medium confidence |
| Vals GPQA Diamond | 95.2% | measured |
| Vals LiveCodeBench | 82.6% | measured |
| Vals MMLU-Pro | 89.1% | measured |
| Vals SWE-bench | 96.2% | measured |
| Terminal-Bench 2.1 (Vals AI) | 85.8% | measured |
| VulcanBench v3 | 87.0% | measured |
| AA AutomationBench | 65.6% | estimated ± 5.5 pp, low confidence |
| GDP.pdf | 24.4% | estimated ± 3.8 pp, high confidence |
| AA ITBench | 56.2% | measured |
| AA Tau3 Banking | 44.1% | estimated ± 3.0 pp, low confidence |
| AA Terminal-Bench 2.1 | 84.9% | estimated ± 3.5 pp, medium confidence |
| AA Terminal-Bench 4.0 | 33.8% | estimated ± 8.9 pp, low confidence |
| CursorBench 4.0 | 41.7% | measured |
| CWE-bench v1 | 59.4% | estimated ± 6.9 pp, medium confidence |
| DeepSWE | 72.7% | measured |
| HLE w/o tools | 46.6% | estimated ± 4.4 pp, high confidence |
| OSWorld 2.0 | 62.6% | measured |
| PostTrainBench v1.1 | 36.2% | measured |
| ProgramBench | 90.9% | estimated ± 4.8 pp, medium confidence |
| Toolathlon-Verified | 79.5% | estimated ± 1.0 pp, medium confidence |
| Vibe Code Bench | 77.7% | estimated ± 10.3 pp, low confidence |
| CharXiv | 86.2% | estimated ± 5.4 pp, medium confidence |
| CharXiv w/o tools | 87.7% | estimated ± 1.8 pp, low confidence |
| ExploitBench | 73.5% | measured |
| ExploitGym | 33.7% | measured |
| C-Eval | 95.6% | estimated ± 0.9 pp, low confidence |
| Claw-Eval | 81.7% | estimated ± 7.9 pp, low confidence |
| HMMT Nov 2025 | 93.2% | estimated ± 2.6 pp, low confidence |
| MathVision | 95.6% | estimated ± 3.6 pp, medium confidence |
| MCP Atlas | 86.3% | estimated ± 3.5 pp, low confidence |
| MCP-Tasks | 80.4% | estimated ± 7.0 pp, low confidence |
| MMAnswerBench | 100.0% | estimated ± 1.3 pp, low confidence |
| MMLU-Redux | 97.8% | estimated ± 1.1 pp, low confidence |
| MMMU-Pro | 83.0% | measured |
| NL2Repo | 60.1% | estimated ± 4.7 pp, high confidence |
| QwenClawBench | 100.0% | estimated ± 6.0 pp, low confidence |
| ScreenSpot Pro | 85.3% | estimated ± 10.0 pp, low confidence |
| SuperGPQA | 88.8% | estimated ± 7.5 pp, low confidence |
| Terminal-Bench 2.0 | 83.2% | estimated ± 5.9 pp, low confidence |
| Toolathlon | 58.0% | measured |
| VideoMMMU | 89.0% | estimated ± 1.0 pp, medium confidence |
| ERQA | 68.8% | estimated ± 3.9 pp, medium confidence |
| HealthBench Hard | 33.1% | measured |
| LiveCodeBench Pro | 87.2% | estimated ± 6.0 pp, low confidence |
| MedXpertQA (MM) | 80.6% | estimated ± 2.4 pp, high confidence |
| MedXpertQA (Text) | 70.7% | estimated ± 4.5 pp, low confidence |
| MMLU-Pro (Arcee) | 88.4% | estimated ± 3.4 pp, low confidence |
| React Native Evals | 100.0% | estimated ± 2.4 pp, low confidence |
| ResearchClawBench | 21.0% | estimated ± 2.0 pp, medium confidence |
| SWE-Rebench | 58.1% | estimated ± 8.9 pp, low confidence |
| ARC-AGI-3 | 7.8% | measured |
| FrontierMath (legacy) | 89.0% | measured |
| BioMysteryBench (human-solvable) | 89.5% | estimated ± 0.6 pp, low confidence |
| DRACO | 84.4% | estimated ± 3.0 pp, medium confidence |
| FrontierCode 1.1 Extended | 60.6% | measured |
| HealthBench (raw) | 57.4% | estimated ± 6.4 pp, low confidence |
| HealthBench (length-adjusted) | 53.9% | estimated ± 0.9 pp, medium confidence |
| HealthBench Professional | 60.5% | measured |
| HealthBench Professional (raw) | 67.2% | estimated ± 5.0 pp, medium confidence |
| HLE-Verified | 54.5% | measured |
| LABBench2 | 82.1% | measured |
| MMMU | 90.3% | estimated ± 0.9 pp, medium confidence |
| LiveCodeBench | 100.0% | estimated ± 8.9 pp, low confidence |
| Agents' Last Exam | 28.8% | estimated ± 2.1 pp, high confidence |
| OpenHarmony Bench | 58.8% | estimated ± 2.4 pp, high confidence |
| SEC-Bench Pro | 71.2% | measured |
| SimpleVQA | 69.8% | estimated ± 8.2 pp, medium confidence |
| MMMLU | 91.2% | estimated ± 1.1 pp, low confidence |
| MMLU | 96.0% | estimated ± 1.1 pp, low confidence |
| MMMU-Pro w/ Python | 84.6% | measured |
| RealWorldQA | 89.3% | estimated ± 1.2 pp, low confidence |
| Video-MME (with subtitle) | 90.0% | estimated ± 1.8 pp, low confidence |
| OmniDocBench 1.5 | 91.7% | estimated ± 5.9 pp, low confidence |