Examples
Worked examples
- Is an instance
A research assistant that retrieves PubMed abstracts via vector search and feeds them to GPT-4 to answer a clinical question
Counter-examples
Looks similar, but isn't
- Not an instance
A plain LLM that answers from its training corpus without retrieving any external documents is not RAG
Editorial commentary
Retrieval-augmented generation (RAG) is an AI architecture in which a language model is supplemented at inference time with documents retrieved from an external corpus — typically via vector-similarity search over an embedding index — so that its output is grounded in retrieved evidence rather than relying solely on knowledge encoded in its parameters during training. A RAG system therefore has two components a plain language model does not: a retrieval index (the corpus it searches) and a retrieval step (what was actually pulled back for a given query), both of which should be disclosed alongside any output.
What RAG does and does not fix
RAG reduces, but does not eliminate, hallucination: a model can still misread, misquote, or over-generalise from retrieved passages, and a poorly configured retrieval step can surface irrelevant or outdated documents that the model then treats as authoritative. Grounding in retrieved text is not the same as verified accuracy.
Disclosure expectations for scholarly use
Where RAG is used for literature search, evidence synthesis, or drafting assistance in research, the corpus searched (which database, what date range, what access restrictions), the retrieval method, and the underlying model version should all be recorded — the retrieved-document set is itself a form of methodology that affects reproducibility of the output, in the same way a search strategy is reported in a systematic review.
References
Also known as
RAG · Retrieval-augmented LLM
Machine-readable encodings
Use in your systems
<role vocab="credit"
vocab-identifier="https://casrai.org/dictionary/"
vocab-term="Retrieval-augmented generation (RAG)"
vocab-term-identifier="https://casrai.org/dictionary/term/retrieval-augmented-generation" />{
"@context": "https://schema.org",
"@type": "DefinedTerm",
"@id": "https://casrai.org/dictionary/term/retrieval-augmented-generation",
"name": "Retrieval-augmented generation (RAG)",
"identifier": "https://casrai.org/dictionary/term/retrieval-augmented-generation",
"description": "An AI architecture in which an LLM is augmented at inference time with documents retrieved from an external corpus (often via vector similarity search), so that the model's outputs are grounded in retrieved evidence rather than relying solely on parametric knowledge.",
"inDefinedTermSet": "https://casrai.org/dictionary/domain/genai-disclosure#set",
"url": "https://casrai.org/dictionary/term/retrieval-augmented-generation",
"sameAs": [
"RAG",
"Retrieval-augmented LLM"
],
"license": "https://creativecommons.org/licenses/by/4.0/",
"publisher": {
"@id": "https://casrai.org/#organization"
},
"author": {
"@id": "https://casrai.org/#editorial-team"
},
"datePublished": "2026-05-21T01:57:21",
"dateModified": "2026-08-22T15:44:00",
"inLanguage": "en-GB",
"isAccessibleForFree": true
}
