Configure WorkBuddy
WorkBuddy officially supports configuring third-party models through its visual interface. Use the settings wizard directly; you do not have to edit a configuration file by hand.
Add InOneAPI
- Open Settings → Models, add a custom model, and select Custom.
- Enter the following parameters.
- Set tool calling, image input, reasoning and the token limit according to the model's actual capabilities in its detail page.
- Save, then select the model in the chat model picker.
| Field | Value |
|---|---|
| Name / provider name | InOneAPI |
| URL | https://api.inoneapi.com/v1/chat/completions |
| API Key | InOneAPI API key |
| Model name / ID | YOUR_MODEL_ID from the model detail page |
| Protocol | Standard OpenAI Chat Completions |
With the standard endpoint, keep the default path handling; there is no need to enable the advanced "Custom protocol" switch. That switch exists for non-standard paths and skips path validation and automatic completion once enabled.
Save and verify
The application stores the configuration locally. The official documentation still supports the legacy ~/.codebuddy/models.json and allows editing it in the interface, so new users do not have to write the old path by hand.
Send a simple task, then check the model, status code and tool calls in InOneAPI Trace. Protocol compatibility does not mean every agent capability is available; treat the model detail page as the source of truth for image and tool support.
CC Switch import is not offered here. CodeBuddy IDE uses a different entry point; see Configure CodeBuddy IDE.
Official reference: WorkBuddy model configuration.