Set up Cherry Studio
Before starting, create a Guangji API key and check account balance and key quota.
1. Add a provider
Open Cherry Studio settings, find the model providers page, and add a provider named Guangji API. Select an OpenAI-compatible provider type. Menu names can vary by version.
2. Enter the key and address
| Setting | Value |
|---|---|
| API Key | The complete key copied from the Guangji console |
| API address | https://guangjiapi.top when the client appends the appropriate path |
If your version requires a full OpenAI Base URL, enter https://guangjiapi.top/v1. Check the complete URL displayed by the client: chat completions should use https://guangjiapi.top/v1/chat/completions.
Use any special path-override syntax only if you understand it. Otherwise, use a normal base URL to avoid duplicate paths.
3. Fetch or add a model
Try fetching the model list. If this is unavailable, manually add the exact model ID from the marketplace.
The model must be available in your key's group. Do not copy model names from an unrelated screenshot.
4. Start a chat
Enable and save the provider. Return to chat, choose a model under Guangji API, and send a short message. After receiving a response, check the model and usage in the Guangji console.
Common issues
- 401: verify the full key and check that it is enabled and unexpired.
- 404: check for duplicate
/v1segments and the correct endpoint type. - Model not found: verify the model ID, group and key restrictions.
- Connection test passes, chat fails: the test may use different parameters or models; inspect the real request error.
See troubleshooting or the official Cherry Studio documentation for client-specific features.