RAG explained — Retrieve material, then write an answer
Retrieval-augmented generation finds relevant material and supplies it to a model before generating an answer, like consulting the right pages in an open-book task.
An everyday example
A company assistant retrieves the current leave policy, answers your question and points to the relevant section.
When you’ll encounter it
RAG is useful for answers grounded in manuals, internal documents or changing information.
What to watch for
Quality depends on sources, retrieval, freshness and permissions. Check that the answer faithfully reflects the retrieved material.
Related terms & sources
Prepared from the public sources below. Examples illustrate the concept. Reviewed on 2026-10-15.
