GraphRAG is a form of RAG in which the AI retrieves knowledge not only as text snippets but as a knowledge graph with relationships.
Illustration: AI-generated · AI Compare Ninja · CC BY 4.0
Example
For the question “Which topics run through all the minutes?”, the graph helps find connections across many documents.
For pros
A language model extracts entities and relationships from the documents, builds clusters and summarises them. Microsoft Research published the approach in 2024. It helps with overview questions but is more expensive to build than plain RAG.
