Zhipu (GLM) · model
GLM-5.3 benchmark scores
As of 2026-10-07, GLM-5.3 (Zhipu (GLM)) has measured scores on 36 benchmarks and estimated scores on 63 more.
| Benchmark | Score | Source |
|---|---|---|
| AA-GPQA Diamond | 91.7% | measured |
| AA-HLE | 42.3% | measured |
| AA-SciCode | 59.0% | measured |
| Artificial Analysis Intelligence Index | 44.8% | measured |
| CritPt | 19.1% | measured |
| GDPval-AA | 57.6% | measured |
| GPQA Diamond | 89.6% | estimated ± 1.7 pp, high confidence |
| HLE | 51.1% | estimated ± 9.5 pp, medium confidence |
| AA-LCR | 79.7% | measured |
| LiveCodeBench v6 | 93.1% | estimated ± 3.1 pp, medium confidence |
| AA-Omniscience Accuracy | 33.9% | measured |
| SciCode | 37.8% | estimated ± 2.9 pp, low confidence |
| Terminal-Bench 2.1 | 88.2% | measured |
| BrowseComp | 85.0% | estimated ± 1.8 pp, medium confidence |
| HLE w/ tools | 62.5% | measured |
| LongBench v2 | 61.5% | estimated ± 5.9 pp, low confidence |
| τ²-bench results | 90.1% | estimated ± 2.4 pp, medium confidence |
| AA Coding Index | 74.8% | measured |
| APEX-Agents | 46.0% | estimated ± 10.0 pp, low confidence |
| APEX-Agents-AA | 38.4% | estimated ± 3.2 pp, medium confidence |
| BioMysteryBench (human-difficult) | 37.5% | estimated ± 5.9 pp, low confidence |
| DeepSearchQA | 93.8% | estimated ± 4.5 pp, high confidence |
| SWE-bench Verified | 82.7% | estimated ± 2.4 pp, high confidence |
| AutomationBench | 48.2% | measured |
| BFCL v4 | 78.0% | estimated ± 2.6 pp, medium confidence |
| CyberGym | 84.5% | measured |
| JobBench | 56.1% | estimated ± 3.4 pp, low confidence |
| SWE-bench Pro | 62.4% | estimated ± 3.6 pp, high confidence |
| WideResearch | 85.5% | estimated ± 3.4 pp, low confidence |
| AA Agentic Index | 53.4% | measured |
| MMLU-Pro | 85.7% | estimated ± 3.5 pp, high confidence |
| GPQA | 90.6% | estimated ± 1.6 pp, high confidence |
| Gert Labs | 69.6% | estimated ± 6.0 pp, low confidence |
| AA-AnalystAgent | 55.1% | estimated ± 2.5 pp, medium confidence |
| AA EnterpriseOps-Gym | 36.4% | measured |
| AA Harvey LAB | 93.6% | estimated ± 1.2 pp, medium confidence |
| ApprenticeBench | 25.9% | estimated ± 4.0 pp, medium confidence |
| ARC-AGI-1 | 94.1% | estimated ± 5.4 pp, medium confidence |
| ARC-AGI-2 | 52.7% | estimated ± 12.6 pp, low confidence |
| CursorBench 3.1 | 66.2% | estimated ± 3.4 pp, medium confidence |
| CursorBench 3.2 | 62.7% | estimated ± 3.3 pp, medium confidence |
| Terminal-Bench 3.0 | 15.0% | estimated ± 6.9 pp, medium confidence |
| FrontierCode 1.1 Main | 45.6% | estimated ± 3.7 pp, low confidence |
| FrontierSWE v2 | 30.2% | measured |
| OSWorld-Verified | 80.0% | estimated ± 4.7 pp, high confidence |
| Vals GPQA Diamond | 88.1% | measured |
| Vals LiveCodeBench | 80.5% | measured |
| Vals MMLU-Pro | 86.8% | measured |
| Vals SWE-bench | 95.4% | measured |
| Terminal-Bench 2.1 (Vals AI) | 71.5% | measured |
| VulcanBench v3 | 78.3% | measured |
| AA AutomationBench | 62.2% | measured |
| GDP.pdf | 11.2% | measured |
| AA ITBench | 46.1% | measured |
| AA Tau3 Banking | 50.3% | measured |
| AA Terminal-Bench 2.1 | 83.9% | measured |
| AA Terminal-Bench 4.0 | 41.9% | measured |
| CursorBench 4.0 | 39.0% | estimated ± 1.6 pp, high confidence |
| CWE-bench v1 | 58.1% | estimated ± 6.0 pp, medium confidence |
| DeepSWE | 66.9% | measured |
| HLE w/o tools | 37.0% | estimated ± 5.1 pp, medium confidence |
| OSWorld 2.0 | 19.4% | estimated ± 5.5 pp, low confidence |
| PostTrainBench v1.1 | 36.2% | estimated ± 2.7 pp, high confidence |
| ProgramBench | 19.0% | measured |
| Terminal-Bench 4.0 | 45.5% | estimated ± 5.1 pp, medium confidence |
| Terminal-Bench-Science 0.1 | 34.9% | estimated ± 4.1 pp, low confidence |
| Toolathlon-Verified | 73.0% | measured |
| Vibe Code Bench | 67.0% | estimated ± 10.3 pp, low confidence |
| ExploitBench | 54.4% | measured |
| ExploitGym | 15.0% | measured |
| C-Eval | 96.1% | estimated ± 1.0 pp, low confidence |
| Claw-Eval | 85.3% | estimated ± 13.6 pp, low confidence |
| MCP Atlas | 85.9% | estimated ± 2.7 pp, low confidence |
| MMLU-Redux | 96.0% | estimated ± 1.1 pp, medium confidence |
| NL2Repo | 58.0% | measured |
| QwenClawBench | 100.0% | estimated ± 6.0 pp, low confidence |
| SuperGPQA | 77.5% | estimated ± 7.5 pp, medium confidence |
| Terminal-Bench 2.0 | 74.5% | estimated ± 5.9 pp, medium confidence |
| Toolathlon | 55.9% | estimated ± 6.3 pp, low confidence |
| HealthBench Hard | 34.4% | estimated ± 7.7 pp, low confidence |
| MMLU-Pro (Arcee) | 100.0% | estimated ± 4.0 pp, low confidence |
| React Native Evals | 100.0% | estimated ± 2.4 pp, low confidence |
| ResearchClawBench | 21.2% | estimated ± 2.0 pp, medium confidence |
| SWE-Rebench | 58.1% | estimated ± 8.9 pp, low confidence |
| ARC-AGI-3 | 1.7% | estimated ± 3.4 pp, medium confidence |
| BioMysteryBench (human-solvable) | 89.0% | estimated ± 0.6 pp, medium confidence |
| DRACO | 85.2% | estimated ± 3.0 pp, medium confidence |
| FrontierCode 1.1 Extended | 59.2% | estimated ± 2.4 pp, medium confidence |
| HealthBench (raw) | 51.2% | estimated ± 6.4 pp, low confidence |
| HealthBench (length-adjusted) | 54.0% | estimated ± 3.7 pp, medium confidence |
| HealthBench Professional | 57.9% | estimated ± 1.3 pp, medium confidence |
| HealthBench Professional (raw) | 61.4% | estimated ± 5.0 pp, medium confidence |
| HLE-Verified | 42.3% | estimated ± 2.0 pp, medium confidence |
| LABBench2 | 81.3% | estimated ± 1.8 pp, low confidence |
| LiveCodeBench | 100.0% | estimated ± 8.9 pp, low confidence |
| Agents' Last Exam | 28.5% | measured |
| OpenHarmony Bench | 60.8% | measured |
| SEC-Bench Pro | 67.7% | estimated ± 4.3 pp, medium confidence |
| MMLU | 92.0% | estimated ± 1.1 pp, medium confidence |