RAG (Retrieval-Augmented Generation)
A technique in which a model retrieves relevant documents at answer time and uses them to generate a grounded response, often with citations.
Short answer
What is RAG?
A technique in which a model retrieves relevant documents at answer time and uses them to generate a grounded response, often with citations.
Why it matters
RAG (Retrieval-Augmented Generation) is part of how search is shifting from ranked links to understood, generated answers. Knowing the term helps you decide where to invest in visibility.
Related terms & services
See how AI search understands your business
Get an AI Search Visibility Audit covering Google, Google AI features, ChatGPT, Gemini and Perplexity – with a prioritised action plan.