Configuring Raccoon
SenseTime Raccoon is a desktop AI workspace that supports OpenAI-compatible custom model services for local chat and agent tasks.
1. Add a model
Open the model menu in chat and choose + Add custom model, or go to Settings → Capabilities & integrations → Model services. Under Custom Models, select + New preset.
| Field | Value |
|---|---|
| Provider | OpenAI-compatible |
| Base URL | https://api.wylon.cn/v1 |
| API Key | Your wylon API key |
| Model name | A complete model ID, such as ZhipuAI/GLM-5.3 |
warning
A wylon connection requires an API keyThe note that built-in services may leave this field empty applies only to Raccoon's own services. wylon is an external service and returns 401 when the key is missing.
Base URL
The base URL should stop at /v1.
Model name
The model-name field is sent directly as the API's model value, so use the exact ID from the Model Plaza or GET /v1/models.
2. Select and verify the model
- Return to chat and open the model menu.
- Select the new custom model.
- Send a simple message. A normal response confirms the configuration.
3. Manage custom models
Go to Settings → Capabilities & integrations → Model services.
| Action | How it works |
|---|---|
| Rename | Select the pencil icon and replace the default “Untitled” name |
| Delete | Select the trash icon to remove the preset |
| Switch | Select a card to edit that preset |
save
Save after every changeSwitching, renaming, and deleting take effect only after you select Save.
4. Use another model
Open the Model Plaza, or request GET https://api.wylon.cn/v1/models. Add or edit a preset, keep the same base URL and API key, and replace the model name with another complete model ID.
info
Create a new task after switching models. For tool use, long-context, or structured-output workloads, rerun the relevant capability checks and adjust limits to the new model.
