WorkBuddy API Key Setup: Run DeepSeek V4 and Other Custom Models in Tencent's Agent

WorkBuddy’s desktop app supports custom AI models through provider API keys, allowing users to add OpenAI-compatible services such as Atlas Cloud and access models including DeepSeek V4 Pro that are absent from its built-in menu. WorkBuddy account API keys are intended for the CodeBuddy CLI, whereas the custom-model configuration requires a separate model-provider key and is available only in the desktop app’s Settings and Models area, not the web app. In testing, DeepSeek V4 Pro processed a complex expense file by creating and running a Python script, producing cleaned output and validating totals, while custom-model use did not consume WorkBuddy credits because provider token charges applied instead. The configuration requires the provider endpoint, exact model ID, API key, and enabled tool calling, although a successful connection test alone does not verify tool-use capability; an API request that returns a tool-calls finish reason can confirm this. The material also notes available built-in, subscription, and free model options, common setup errors such as incorrect endpoints or display names, and warns that locally stored API keys may be saved in plain text.

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.