Zhipu (GLM) · model
GLM-5.2 benchmark scores
As of 2026-11-22, GLM-5.2 (Zhipu (GLM)) has measured scores on 53 benchmarks and estimated scores on 115 more.
| Benchmark | Score | Source |
|---|---|---|
| AA-GPQA Diamond | 89.5% | measured |
| AA-HLE | 41.1% | measured |
| AA-SciCode | 51.2% | measured |
| AIME26 | 99.2% | measured |
| Artificial Analysis Intelligence Index | 33.7% | measured |
| CritPt | 20.9% | measured |
| GDPval-AA | 43.6% | measured |
| GPQA Diamond | 91.2% | measured |
| HLE | 40.5% | measured |
| IFBench | 71.3% | estimated ± 6.6 pp, medium confidence |
| AA-LCR | 78.3% | measured |
| LiveCodeBench v6 | 91.6% | estimated ± 1.7 pp, high confidence |
| AA-Omniscience Accuracy | 24.3% | measured |
| SciCode | 51.0% | estimated ± 3.1 pp, high confidence |
| Terminal-Bench 2.1 | 81.0% | measured |
| BrowseComp | 85.3% | estimated ± 4.2 pp, high confidence |
| HLE w/ tools | 54.7% | measured |
| IFEval | 92.9% | estimated ± 2.8 pp, high confidence |
| LongBench v2 | 64.2% | estimated ± 5.5 pp, medium confidence |
| VITA-Bench | 46.9% | estimated ± 9.9 pp, low confidence |
| τ²-bench results | 99.1% | measured |
| AA Coding Index | 68.8% | measured |
| APEX-Agents | 34.3% | estimated ± 4.9 pp, low confidence |
| APEX-Agents-AA | 33.7% | measured |
| BioMysteryBench (human-difficult) | 48.3% | estimated ± 6.8 pp, low confidence |
| DeepSearchQA | 91.6% | estimated ± 4.5 pp, high confidence |
| SWE-bench Verified | 80.7% | estimated ± 2.3 pp, high confidence |
| BFCL v4 | 76.1% | estimated ± 2.7 pp, medium confidence |
| CyberGym | 75.8% | estimated ± 5.3 pp, medium confidence |
| JobBench | 35.3% | estimated ± 7.3 pp, low confidence |
| SkillsBench | 56.4% | estimated ± 9.1 pp, low confidence |
| SWE-bench Pro | 62.1% | measured |
| WideResearch | 80.2% | estimated ± 3.3 pp, high confidence |
| AA Agentic Index | 39.4% | measured |
| MMLU-Pro | 85.6% | estimated ± 2.5 pp, low confidence |
| AA-IFBench | 73.3% | measured |
| FailureSensorIQ | 66.9% | estimated ± 1.7 pp, high confidence |
| GPQA | 91.2% | measured |
| MMLU | 90.5% | estimated ± 0.9 pp, high confidence |
| Vals GPQA Diamond | 85.6% | measured |
| Vals LiveCodeBench | 69.5% | measured |
| Vals MMLU-Pro | 86.7% | measured |
| MATH-500 | 98.2% | estimated ± 3.1 pp, low confidence |
| MMMLU | 90.9% | estimated ± 0.7 pp, low confidence |
| Chess Puzzles | 22.0% | measured |
| AIME 2025 | 100.0% | estimated ± 1.8 pp, low confidence |
| AA MMLU-Pro | 85.0% | estimated ± 0.7 pp, high confidence |
| Gert Labs | 49.1% | measured |
| AA-AnalystAgent | 46.5% | estimated ± 3.7 pp, medium confidence |
| AA EnterpriseOps-Gym | 42.7% | measured |
| AA Harvey LAB v1.0 | 92.0% | estimated ± 1.2 pp, medium confidence |
| AA Tau3 Banking | 39.6% | estimated ± 2.0 pp, medium confidence |
| AA Terminal-Bench 2.1 | 82.1% | estimated ± 2.5 pp, high confidence |
| ApprenticeBench | 10.0% | estimated ± 5.0 pp, low confidence |
| ARC-AGI-1 | 91.1% | estimated ± 2.5 pp, high confidence |
| ARC-AGI-2 | 63.5% | estimated ± 9.7 pp, medium confidence |
| BabaIsBench | 96.0% | measured |
| CursorBench 3.1 | 47.0% | estimated ± 3.4 pp, low confidence |
| CursorBench 3.2 | 55.0% | measured |
| EBR-bench | 9.5% | measured |
| Terminal-Bench 3.0 | 4.6% | measured |
| FrontierCode 1.1 Main | 37.9% | estimated ± 1.6 pp, low confidence |
| FrontierMath v2 (Tier 4) | 29.3% | measured |
| FrontierMath v2 (Tiers 1-3) | 59.2% | measured |
| FrontierSWE v2 | 19.2% | estimated ± 8.3 pp, medium confidence |
| Furniture Assembly | 24.0% | estimated ± 7.4 pp, medium confidence |
| MaxxBench | 33.1% | estimated ± 4.3 pp, low confidence |
| MazeBench | 0.0% | measured |
| MirrorCode | 36.7% | measured |
| Mystery Game Puzzles | 15.0% | measured |
| NYT Connections | 74.3% | measured |
| OSWorld-Verified | 77.7% | estimated ± 5.1 pp, medium confidence |
| Terminal-Bench 4.1 | 0.0% | estimated ± 4.8 pp, medium confidence |
| Terminal-Bench Hard | 50.8% | measured |
| Vals SWE-bench | 82.8% | measured |
| Terminal-Bench 2.1 (Vals AI) | 67.8% | measured |
| VulcanBench v3 | 76.7% | estimated ± 5.1 pp, low confidence |
| CWE-Bench-AA | 55.8% | estimated ± 5.2 pp, medium confidence |
| AA Cyber Index | 24.2% | estimated ± 9.2 pp, low confidence |
| DeepsecBench-AA | 16.9% | estimated ± 8.2 pp, medium confidence |
| HealthBench Professional (length-adjusted, Artificial Analysis) | 56.1% | estimated ± 4.3 pp, low confidence |
| HealthBench Professional (raw, Artificial Analysis) | 66.0% | estimated ± 0.8 pp, medium confidence |
| AA ITBench | 42.7% | measured |
| Bug Hunt Bench | 24.6% | estimated ± 6.3 pp, low confidence |
| CursorBench 4.0 | 30.3% | estimated ± 2.0 pp, medium confidence |
| CWE-bench v1 | 57.4% | estimated ± 5.1 pp, medium confidence |
| DeepSWE | 72.9% | estimated ± 3.8 pp, medium confidence |
| DRACO | 86.9% | estimated ± 0.7 pp, low confidence |
| FrontierCode 2.0 | 16.9% | estimated ± 5.4 pp, low confidence |
| FrontierPhysics | 2.5% | estimated ± 5.0 pp, low confidence |
| HLE w/o tools | 40.5% | measured |
| OSWorld 2.0 | 16.1% | estimated ± 6.9 pp, low confidence |
| OSWorld 2.1 | 61.1% | estimated ± 3.6 pp, high confidence |
| OSWorld 2.1 Offline | 40.2% | estimated ± 1.9 pp, low confidence |
| PostTrainBench v1.1 | 31.7% | measured |
| ProgramBench | 63.7% | measured |
| SpireClimb | 5.8% | estimated ± 7.2 pp, medium confidence |
| StealthMark | 48.0% | estimated ± 8.7 pp, low confidence |
| Toolathlon-Verified | 67.6% | estimated ± 2.3 pp, medium confidence |
| VulcanBench Frontier v4 | 88.5% | estimated ± 1.2 pp, medium confidence |
| AA AutomationBench | 41.9% | estimated ± 7.9 pp, low confidence |
| GDP.pdf | 19.6% | estimated ± 4.2 pp, high confidence |
| AA Terminal-Bench 4.0 | 4.3% | estimated ± 11.0 pp, low confidence |
| ARC-AGI-3 | 3.5% | estimated ± 7.2 pp, medium confidence |
| ARC-AGI-3 Provider Adapter | 2.7% | estimated ± 7.3 pp, medium confidence |
| cursorBench41 | 44.8% | estimated ± 3.7 pp, medium confidence |
| CWE-bench | 46.4% | estimated ± 6.2 pp, medium confidence |
| MKA | 35.5% | estimated ± 4.6 pp, high confidence |
| SkillsBench official leaderboard (BenchFlow) | 51.5% | estimated ± 6.1 pp, medium confidence |
| BinaryAudit | 35.8% | estimated ± 12.1 pp, low confidence |
| CompileBench | 73.3% | estimated ± 10.0 pp, low confidence |
| OTelBench | 17.5% | estimated ± 4.1 pp, low confidence |
| Vibe Code Bench | 50.5% | estimated ± 10.0 pp, low confidence |
| FACTS Multimodal | 42.1% | estimated ± 4.0 pp, high confidence |
| Hermes Bench | 70.0% | estimated ± 1.8 pp, medium confidence |
| Hermes Index | 37.2% | estimated ± 3.1 pp, high confidence |
| ExploitBench | 28.8% | estimated ± 7.2 pp, low confidence |
| ExploitGym | 11.2% | estimated ± 2.7 pp, high confidence |
| C-Eval | 95.7% | estimated ± 1.0 pp, low confidence |
| Claw-Eval | 73.3% | estimated ± 3.2 pp, medium confidence |
| DeepPlanning | 48.5% | estimated ± 13.1 pp, low confidence |
| HMMT Feb 2025 | 96.3% | estimated ± 1.9 pp, high confidence |
| HMMT Feb 2026 | 92.5% | measured |
| HMMT Nov 2025 | 94.4% | measured |
| MCP Atlas | 76.8% | measured |
| MCP-Tasks | 75.6% | estimated ± 7.0 pp, low confidence |
| MMAnswerBench | 91.0% | measured |
| MMLU-Redux | 94.0% | estimated ± 0.8 pp, medium confidence |
| NL2Repo | 48.9% | measured |
| PostTrainBench v1 | 34.3% | measured |
| QwenClawBench | 63.6% | estimated ± 2.6 pp, medium confidence |
| SuperGPQA | 80.3% | estimated ± 7.2 pp, medium confidence |
| Terminal-Bench 2.0 | 67.8% | estimated ± 5.4 pp, medium confidence |
| Toolathlon | 48.2% | measured |
| AA LiveCodeBench | 93.3% | estimated ± 2.6 pp, medium confidence |
| HealthBench Hard | 14.3% | estimated ± 5.1 pp, low confidence |
| LiveCodeBench Pro | 85.3% | estimated ± 6.0 pp, low confidence |
| MedXpertQA (Text) | 57.5% | estimated ± 4.5 pp, medium confidence |
| MMLU-Pro (Arcee) | 87.3% | estimated ± 3.4 pp, low confidence |
| React Native Evals | 81.4% | estimated ± 3.2 pp, high confidence |
| ResearchClawBench | 20.7% | measured |
| PostTrainBench | 30.4% | measured |
| DAYJOB | 1.3% | estimated ± 3.2 pp, high confidence |
| DAYJOB Healthcare | 0.7% | estimated ± 4.2 pp, high confidence |
| Finance Agent v2 | 55.7% | estimated ± 2.4 pp, medium confidence |
| BioMysteryBench (human-solvable) | 89.2% | estimated ± 0.6 pp, medium confidence |
| FrontierCode 1.1 Extended | 55.1% | estimated ± 1.6 pp, low confidence |
| HealthBench (raw) | 49.7% | estimated ± 6.0 pp, medium confidence |
| HealthBench (length-adjusted) | 54.0% | estimated ± 3.4 pp, medium confidence |
| HealthBench Professional | 50.1% | estimated ± 4.2 pp, medium confidence |
| HealthBench Professional (raw) | 60.3% | estimated ± 4.5 pp, high confidence |
| HLE-Verified | 16.4% | estimated ± 2.0 pp, low confidence |
| LABBench2 | 81.3% | estimated ± 1.8 pp, low confidence |
| ScarfBench | 62.9% | estimated ± 5.5 pp, low confidence |
| BlenderWorld (text-only) | 70.0% | estimated ± 2.8 pp, medium confidence |
| Cookie-Bench | 76.9% | estimated ± 1.3 pp, medium confidence |
| STEMBench | 2.4% | estimated ± 0.2 pp, low confidence |
| SimpleQA | 24.0% | estimated ± 4.3 pp, medium confidence |
| Agents' Last Exam | 31.3% | estimated ± 9.7 pp, medium confidence |
| OpenHarmony Bench v1.0 | 58.4% | measured |
| SEC-Bench Pro | 21.1% | estimated ± 7.5 pp, low confidence |
| IMOAnswerBench | 84.6% | estimated ± 5.6 pp, medium confidence |
| MRCRv2 | 85.5% | estimated ± 11.3 pp, low confidence |
| MMLongBench-Doc | 61.3% | estimated ± 2.6 pp, low confidence |
| SWE-Marathon | 56.0% | estimated ± 12.9 pp, low confidence |
| BioBasicsBench | 25.4% | estimated ± 2.2 pp, low confidence |
| LifeVal | 75.2% | estimated ± 3.8 pp, medium confidence |
| ExploitBench Internal Port | 0.0% | estimated ± 1.4 pp, low confidence |