General setup
Connect an OpenAI-compatible client using an API base URL, API key, and model ID.
Before you begin
- Choose a model and copy its ID
- Confirm its group
- Create an API key for that group
- Use a client that supports a custom OpenAI-compatible endpoint
Configuration fields
| Field | Value |
|---|---|
| Provider | OpenAI-compatible or custom OpenAI |
| API base URL | https://milorouter.com/v1 |
| API key | A key created in MiloRouter |
| Model | The exact ID from the model plaza |
Verify the connection
- 1
Save the configuration
Recheck the URL, key, and model ID.
- 2
Send a short request
Keep the first test small.
- 3
Check the response
A normal model reply confirms the basic connection.
- 4
Review usage
Confirm model, tokens, and cost in MiloRouter.
If it does not connect
- Confirm the base URL includes /v1
- Remove spaces around the API key
- Match the model ID exactly
- Confirm the key group
- Check balance and channel status