learn
RAG
What RAG means, how it works, and which tools and models on this site relate to it.
Definition
Retrieval-augmented generation (RAG) is a pattern where a model answers using documents fetched at request time, instead of relying only on weights learned during training.
How it works
A query is embedded, similar chunks are retrieved from a vector index or search system, and those chunks are passed to the model as context. Quality depends on chunking, retrieval, and how faithfully the model cites that context.
Where it shows up
- Internal knowledge assistants
- Customer support over a help center
- Research Q&A over papers