Chapter 9 — Connecting External APIs

You may not have credits, but you have your own LLM API.

WorkBuddy supports connecting other LLM APIs, plus Coding Plan, Token Plan and other packages.

Go in directly from Settings.

Pick the model option.

Click Add Model.

You can pick various coding plans or a custom API.

For example, for DeepSeek you only need to enter an API key.

Or connect a local Ollama model — first start Ollama locally (default port 11434, OpenAI-compatible interface). The advantage of a local model is data never leaves your machine, it can run offline, and there's zero token cost.