Project GraphRAG
LLM-Derived Knowledge Graphs
GraphRAG
GraphRAG (Graphs + Retrieval Augmented Generation) is a technique for richly understanding text datasets by combining text extraction, network analysis, and LLM prompting and summarization into a single end-to-end system.
-
GraphRAG: Unlocking LLM discovery on narrative private data
2024年2月13日 | Jonathan Larson, Steven Truitt
-
GraphRAG: New tool for complex data discovery now on GitHub
2024年7月2日 | Darren Edge, Ha Trinh, Steven Truitt, Jonathan Larson
-
GraphRAG: Improving global search via dynamic community selection
2024年11月15日 | Bryan Li, Ha Trinh, Darren Edge, Jonathan Larson
-
LazyGraphRAG: Setting a new standard for quality and cost
2024年11月25日 | Darren Edge, Ha Trinh, Jonathan Larson
-
Moving to GraphRAG 1.0 – Streamlining ergonomics for developers and users
2024年12月16日 | Nathan Evans, Alonso Guevara Fernández, Joshua Bradley
-
Claimify: Extracting high-quality claims from language model outputs
2025年3月19日 | Dasha Metropolitansky
-
BenchmarkQED: Automated benchmarking of RAG systems
2025年6月5日 | Darren Edge, Ha Trinh, Andres Morales Esquivel, Jonathan Larson
Microsoft Discovery
GraphRAG and LazyGraphRAG technology is now available through Microsoft Discovery (opens in new tab), an agentic platform for scientific research built in Azure.
Open Source
The GraphRAG library is available on GitHub (opens in new tab). The GraphRAG solution accelerator repository, an API experience hosted on Azure, has been archived but can still be referenced below. BenchmarkQED, a set of tools for evaluating RAG based systems, is also available on GitHub (opens in new tab).
GraphRAG Library (opens in new tab)