Retrieval-Augmented Generation (RAG)

Also known as: RAG

Retrieval-augmented generation is an architecture in which relevant documents are retrieved from an external store (usually via vector or hybrid search) and supplied to a language model as context, so responses are grounded in specific source material.

RAG addresses two limits at once: models have a training cutoff, and they cannot know your private data. Retrieval supplies both without retraining.

System quality is usually bounded by retrieval, not generation. If the right document is not retrieved, no model can produce the right answer.