Memoh
Connect Memoh to Bamboo Relay.
Memoh provides self-hosted personal agents with extensible model providers. Bamboo maintains this compatibility guide; the application is maintained by its own project.
Configuration
Create a key in API Keys, and copy an available model ID from Models.
| Setting | Value |
|---|---|
| Provider | OpenAI-compatible / custom OpenAI |
| Base URL | https://api.bamboonode.cn/v1 |
| API key | Your Bamboo inference key |
| Model | Exact available model ID |
Setup
- After deploying Memoh, open model or provider settings.
- Add an OpenAI-compatible provider named
Bamboo Relay. - Enter the Base URL, key and chat model ID.
- For a knowledge base, configure an available Embeddings model and its vector dimensions separately.
- Save and run a conversation or agent task.
Application notes
The server or container must be able to reach api.bamboonode.cn; browser connectivity alone is insufficient. Store the key in deployment secrets or environment variables, not an image or public Compose file. If your version appends /v1, configure the origin instead.
Troubleshooting
401: confirm that the key is complete and enabled. Do not addBearerin a key input field.404: check whether the application appends/v1and avoid a duplicated suffix.- A model or permission error: check the exact ID, supported protocol and API-key group.