How-To Guide
Configure z.ai/GLM as Your Hermes Provider
Configure Hermes Agent to use ZhipuAI's GLM models via z.ai — powerful Chinese language models.
Quick answer
Configure z.ai/GLM (ZhipuAI) for strong reasoning and coding — GLM-4-Plus and GLM-5 are among the more capable models in their class. Add your z.ai/GLM API key in config and pick the model; the community runs GLM as a primary model and via coding plans for predictable cost.
ZhipuAI's GLM models are among the most capable Chinese language models. GLM-4-Plus and GLM-5 offer excellent reasoning and coding capabilities.
Managed cloud · API costs included · Skill library · Cancel anytime
Before you start:
- ☑Hermes Agent installed
- ☑z.ai or ZhipuAI account and API key
Steps
- 1
Get your API key
Sign up at z.ai or open.bigmodel.cn and get your API key
- 2
Add to .env
Add GLM_API_KEY=your_key to ~/.hermes/.env
- 3
Set the provider
In config.yaml, set model: provider: zai
- 4
Choose a model
Set model: default: glm-4-plus or glm-5
Pro Tips
- 💡GLM models have strong bilingual capabilities (Chinese/English)
- 💡Use GLM-4-Plus for balanced performance, GLM-5 for maximum capability
- 💡Endpoint auto-detection probes and caches the correct URL
Troubleshooting
❌ Endpoint auto-detection failing
✅ Set GLM_BASE_URL explicitly: https://api.z.ai/api/paas/v4
❌ Rate limited
✅ z.ai has usage limits. Check your account quota at z.ai dashboard.
FAQ
What are GLM models good for?
Reasoning and coding. GLM-4-Plus and GLM-5 from ZhipuAI are among the most capable models in their class, used as primary models in Hermes.
How do I set up GLM in Hermes?
Configure the z.ai/GLM provider with your API key and select GLM-4-Plus or GLM-5. Some users buy a coding plan for predictable monthly cost.
Direct or via an aggregator for GLM?
Direct usually gives better latency. Aggregators add reach but can be slower; pick based on whether you value speed or single-account convenience.