OpenAI · model
GPT-5.2-Codex benchmark scores
As of 2026-10-07, GPT-5.2-Codex (OpenAI) has measured scores on 14 benchmarks and estimated scores on 80 more.
| Benchmark | Score | Source |
|---|---|---|
| AA-GPQA Diamond | 89.9% | measured |
| AA-HLE | 35.7% | measured |
| AA-SciCode | 51.2% | estimated ± 3.8 pp, high confidence |
| Artificial Analysis Intelligence Index | 28.5% | measured |
| CritPt | 8.7% | measured |
| GDPval-AA | 31.7% | estimated ± 6.2 pp, medium confidence |
| GPQA Diamond | 89.5% | estimated ± 2.2 pp, high confidence |
| HLE | 42.4% | estimated ± 9.5 pp, medium confidence |
| IFBench | 76.3% | estimated ± 6.9 pp, medium confidence |
| AA-LCR | 82.3% | measured |
| AA-Omniscience Accuracy | 41.1% | measured |
| SciCode | 52.7% | estimated ± 2.9 pp, low confidence |
| BrowseComp | 76.5% | estimated ± 5.7 pp, medium confidence |
| IFEval | 93.7% | estimated ± 1.3 pp, high confidence |
| LongBench v2 | 61.5% | estimated ± 5.9 pp, low confidence |
| τ²-bench results | 92.1% | measured |
| AA Coding Index | 56.2% | estimated ± 5.8 pp, medium confidence |
| AA-MMMU-Pro | 76.3% | measured |
| APEX-Agents-AA | 27.3% | estimated ± 3.6 pp, medium confidence |
| BioMysteryBench (human-difficult) | 41.5% | estimated ± 5.9 pp, low confidence |
| DeepSearchQA | 78.3% | estimated ± 10.9 pp, low confidence |
| SWE-bench Verified | 78.8% | estimated ± 4.4 pp, high confidence |
| BFCL v4 | 67.5% | estimated ± 9.8 pp, low confidence |
| CyberGym | 47.9% | estimated ± 10.0 pp, low confidence |
| JobBench | 26.0% | measured |
| SWE-bench Pro | 56.9% | estimated ± 2.1 pp, high confidence |
| τ³-bench results | 68.9% | estimated ± 8.9 pp, medium confidence |
| WideResearch | 76.0% | estimated ± 1.2 pp, medium confidence |
| AA Agentic Index | 17.0% | estimated ± 6.4 pp, medium confidence |
| MMLU-Pro | 85.7% | estimated ± 3.5 pp, high confidence |
| GPQA | 89.8% | estimated ± 3.2 pp, high confidence |
| AA-IFBench | 77.6% | measured |
| Gert Labs | 51.8% | measured |
| AA EnterpriseOps-Gym | 41.5% | estimated ± 3.6 pp, medium confidence |
| ApprenticeBench | 0.0% | estimated ± 3.1 pp, low confidence |
| ARC-AGI-1 | 83.6% | estimated ± 5.4 pp, medium confidence |
| ARC-AGI-2 | 46.6% | estimated ± 14.2 pp, low confidence |
| CursorBench 3.1 | 46.3% | estimated ± 5.8 pp, low confidence |
| CursorBench 3.2 | 54.1% | estimated ± 5.4 pp, low confidence |
| FrontierCode 1.1 Main | 17.3% | estimated ± 4.3 pp, low confidence |
| FrontierSWE v2 | 7.1% | estimated ± 9.5 pp, low confidence |
| OfficeQA Pro | 50.3% | estimated ± 7.2 pp, low confidence |
| OSWorld-Verified | 61.9% | estimated ± 5.9 pp, medium confidence |
| Vals GPQA Diamond | 88.5% | estimated ± 7.4 pp, medium confidence |
| Vals LiveCodeBench | 88.0% | measured |
| Vals MMLU-Pro | 87.0% | estimated ± 2.2 pp, high confidence |
| Vals SWE-bench | 72.4% | measured |
| VulcanBench v3 | 85.8% | estimated ± 5.6 pp, low confidence |
| AA AutomationBench | 46.8% | estimated ± 5.0 pp, low confidence |
| GDP.pdf | 17.8% | estimated ± 5.2 pp, low confidence |
| AA ITBench | 47.4% | estimated ± 3.6 pp, medium confidence |
| CursorBench 4.0 | 26.5% | estimated ± 4.4 pp, low confidence |
| DeepSWE | 56.3% | estimated ± 7.2 pp, low confidence |
| HLE w/o tools | 33.4% | estimated ± 5.7 pp, medium confidence |
| OSWorld 2.0 | 1.0% | estimated ± 4.5 pp, low confidence |
| PostTrainBench v1.1 | 19.3% | estimated ± 2.7 pp, high confidence |
| Toolathlon-Verified | 73.2% | estimated ± 1.3 pp, low confidence |
| Vibe Code Bench | 37.9% | measured |
| CharXiv | 83.7% | estimated ± 7.5 pp, low confidence |
| CharXiv w/o tools | 80.4% | estimated ± 2.4 pp, high confidence |
| C-Eval | 93.8% | estimated ± 1.0 pp, low confidence |
| Claw-Eval | 50.7% | estimated ± 5.6 pp, low confidence |
| MathVision | 89.0% | estimated ± 4.9 pp, high confidence |
| MCP Atlas | 49.2% | estimated ± 8.2 pp, medium confidence |
| MCP-Tasks | 68.2% | estimated ± 8.2 pp, low confidence |
| MMLU-Redux | 96.7% | estimated ± 1.1 pp, low confidence |
| MMMU-Pro | 77.5% | estimated ± 4.4 pp, medium confidence |
| NL2Repo | 41.7% | estimated ± 4.7 pp, high confidence |
| ScreenSpot Pro | 75.1% | estimated ± 10.7 pp, low confidence |
| SuperGPQA | 81.7% | estimated ± 7.5 pp, medium confidence |
| Toolathlon | 43.0% | estimated ± 7.2 pp, medium confidence |
| ERQA | 65.7% | estimated ± 3.9 pp, high confidence |
| HealthBench Hard | 24.5% | estimated ± 7.7 pp, low confidence |
| MedXpertQA (MM) | 74.2% | estimated ± 6.5 pp, low confidence |
| MMLU-Pro (Arcee) | 100.0% | estimated ± 4.0 pp, low confidence |
| React Native Evals | 69.2% | estimated ± 2.4 pp, low confidence |
| ResearchClawBench | 16.1% | estimated ± 2.2 pp, medium confidence |
| SWE-Rebench | 61.0% | estimated ± 6.5 pp, low confidence |
| ARC-AGI-3 | 1.7% | estimated ± 3.4 pp, medium confidence |
| BioMysteryBench (human-solvable) | 82.5% | estimated ± 0.6 pp, low confidence |
| FrontierCode 1.1 Extended | 18.8% | estimated ± 2.4 pp, low confidence |
| HealthBench (raw) | 45.0% | estimated ± 6.4 pp, low confidence |
| HealthBench (length-adjusted) | 53.9% | estimated ± 3.7 pp, low confidence |
| HealthBench Professional | 57.7% | estimated ± 4.8 pp, low confidence |
| HealthBench Professional (raw) | 55.6% | estimated ± 5.0 pp, low confidence |
| HLE-Verified | 17.6% | estimated ± 2.0 pp, low confidence |
| LABBench2 | 76.9% | estimated ± 1.8 pp, low confidence |
| MMMU | 84.1% | estimated ± 0.9 pp, high confidence |
| Agents' Last Exam | 48.9% | estimated ± 8.7 pp, low confidence |
| OpenHarmony Bench | 51.9% | estimated ± 3.2 pp, high confidence |
| MMLU | 93.6% | estimated ± 1.1 pp, low confidence |
| MMMU-Pro w/ Python | 79.5% | estimated ± 1.6 pp, high confidence |
| RealWorldQA | 85.7% | estimated ± 1.2 pp, medium confidence |
| OmniDocBench 1.5 | 91.7% | estimated ± 6.4 pp, low confidence |