AI & LLMs

Retrieval-augmented generation (RAG)

Also known as: RAG

RAG fetches relevant documents at query time and feeds them to a language model as context, so answers are grounded in your data instead of only the model's training.

Retrieval-augmented generation (RAG) combines a retriever with a generator. When a question comes in, the system first retrieves relevant passages - usually via vector search over embeddings - then passes them to a language model as context so its answer is grounded in those sources.

RAG is the common alternative to fine-tuning when you need a model to answer over private, large, or frequently changing data without retraining it.

In GraphCanon

Retrieval, vector database, and framework tools that implement RAG pipelines are a core slice of the graph. Compare pages and stacks show how these pieces (embeddings, a vector store, an orchestration framework) fit together.

See also

Related terms

Last reviewed 2026-07-09

Command menu

Search tools or jump to a page