Definition
Retrieval-Augmented Generation (RAG)
Retrieval-augmented generation (RAG) is a technique where an AI system retrieves relevant documents from an external source at query time and generates its answer from that fetched context rather than from model weights alone. It is the mechanism behind AI search: the model reads live pages before it writes.
RAG is why fresh, crawlable, clearly-structured pages can influence answers without waiting for a model retraining cycle — if your page is retrieved, it shapes the response. For GEO, RAG is the lever that pays off in weeks, versus training-data memory that shifts over months.
Related terms
Reading definitions won’t tell you where you stand
Run the free check and see whether AI actually recommends your brand.
Check my AI visibility