GLM 5.2
Model ID: vllm/release/glm-5-2 Z.AI (Zhipu)
Controlling thinking mode: To enable or disable thinking for GLM, use the chat_template_kwargs field: set chat_template_kwargs.enable_thinking to true or false. A top-level enable_thinking field is not recognized by the runtime.
Reasoning effort: GLM supports reasoning_effort via chat_template_kwargs.reasoning_effort with values "high" and "max". Use "max" for complex tasks, "high" for default thinking.
Risk Assessment (EU AI Act)
Section titled “Risk Assessment (EU AI Act)”| Risk Factor | Category | Level |
|---|---|---|
| Model size | Deepfake potential | Medium |
| Open source (MIT) | Misuse potential | Low |
| Environmental impact | Model size (753B) | Medium-high |
| Deprecated | End of Life | 2026-10-31 |
Training Data & Sustainability
Section titled “Training Data & Sustainability”Training data: publicly available in the technical report (arXiv: 2602.15763).
Sustainability:
| Metric | Value |
|---|---|
| CO₂ | High (753B model), specific figures not published |
| Optimization | IndexShare reduces FLOPs by 2.9× on long contexts |
Original Sources
Section titled “Original Sources”- HuggingFace: https://huggingface.co/zai-org/GLM-5.2
- License: MIT License
- Technical report: https://arxiv.org/abs/2602.15763
Model card created on July 2, 2026. This model card serves EU AI Act transparency purposes (Regulation (EU) 2024/1689). The information is based on publicly available data from the model provider.
