WorkBuddy
Enter this site and your key in WorkBuddy custom model settings.
WorkBuddy is Tencent's desktop assistant for office work and coding. Add this site as a custom model; it only supports the OpenAI-compatible format.
- Protocol: OpenAI-compatible (Chat Completions)
- Base URL:
https://tokenfunc.com/v1 - Models: any, for example
deepseek-chat
Open the model settings
Click your avatar at the bottom left, go to Settings → Models and click Add Model.
Screenshot pendingThe Models page in WorkBuddy settings with the Add Model buttontools/workbuddy-models.png
Choose the custom provider and fill it in
Choose Custom as the provider and fill in:
Field Value URL https://tokenfunc.com/v1API Key sk-xxxxxxxxxxxxxxxxModel name deepseek-chatWorkBuddy appends
/chat/completionsitself, so stop at/v1here.Screenshot pendingThe custom model dialog with URL, API Key and model name filled intools/workbuddy-custom.png
Set the model capabilities if needed
Under Advanced, tick "Supports tool calls", "Supports images" and so on to match the model. Keep the defaults if unsure.
If you turn on the Custom protocol switch in the advanced settings, the URL must be the full path:
https://tokenfunc.com/v1/chat/completions.Save and verify
After saving, open the model picker in the chat box, find the custom model group, select the new model and send a message.
On first use WorkBuddy may show a relay service risk warning. That is its notice for non-official addresses; confirm the address is
https://tokenfunc.comand continue.
Adding many models with the config file
To add several models at once, edit ~/.codebuddy/models.json (Windows: %USERPROFILE%\.codebuddy\models.json). Changes apply without a restart. Note that url here must be the full path:
{
"models": [
{
"id": "deepseek-chat",
"name": "tokenfunc deepseek-chat",
"vendor": "OpenAI",
"apiKey": "sk-xxxxxxxxxxxxxxxx",
"url": "https://tokenfunc.com/v1/chat/completions",
"maxInputTokens": 200000,
"maxOutputTokens": 8192,
"supportsToolCall": true,
"supportsImages": false
}
]
}Save the file as UTF-8 without a BOM, or it may fail to load.
Troubleshooting
| Symptom | Cause and fix |
|---|---|
| 404 | The URL in the UI is missing /v1; or Custom protocol is on without the full path; or url in models.json stops at /v1 |
| 401 | Wrong key or no quota left. Check it on API Keys |
| No Custom provider | Your enterprise admin disabled custom models. Ask the admin |
| Model does not show up after adding | The model ID clashes with a built-in model. Use another name or update to the latest version |
| 400 when sending an image | The model cannot read images, or "Supports images" is not ticked |