RAG (Retrieval-Augmented Generation)
Retrieval-Augmented Generation: generating an answer using information retrieved from an external source, such as a document collection, database or web search.
Explanation
A RAG system retrieves relevant information and supplies it as context to a language model. For example, a support assistant can retrieve a delivery policy before answering a delivery question.
Systems may retrieve documents or smaller passages. A displayed citation can link to the original page even when only one passage informed the answer. Retrieval and citation are separate steps; retrieved information need not appear in the final answer.
Clear headings and self-contained explanations help people understand individual passages. They do not guarantee retrieval, citation or a correct answer. RAG describes a general approach, not one identical implementation across AI services.
Sources
Updated: Oct 1, 2026