llm

Overview ¶

Package llm generates answers, through either a local Ollama model or a user-supplied API key.

A free path must always exist: someone who will not pay for a key still gets working answers. With neither Ollama nor a key configured, retrieval still returns ranked passages — what is lost is the prose, not the results.