GraphRAG is Microsoft Research’s MIT-licensed Python pipeline that builds an LLM-extracted entity knowledge graph over a text corpus and answers queries from community summaries, the reference implementation of the graph-based RAG pattern.
It founded the graph-RAG family and answered the question vector RAG answers worst (what are the themes across this whole corpus), and its own README now declares the project largely in maintenance mode, which makes it a pattern to study rather than a dependency to adopt.
What it is #
A pip install graphrag CLI and library that chunks a corpus, prompts an LLM to extract entities, relationships, and claims per chunk, detects communities over the graph, and pre-generates community summaries for global questions.
It answers two query styles: local search (entity-anchored neighborhood questions) and global search (map-reduce over community summaries).
The method paper (arXiv 2404.16130) frames it as moving from local to global sense-making on narrative private data.
Made by Microsoft Research, first released July 2024, optional Azure OpenAI or local model support.
Status #
Active by the commit clock, retired by its own README. 36,325 stars (as of 2026-10-13) and 3,858 forks, pushed 2026-11-09 at the repository level with the default branch’s latest commit still the 2026-11-02 maintenance-workflow action pinning, 67 open issues (GitHub API, as of 2026-11-15).
PyPI shows version 3.3.0 released 2026-10-08 across 52 releases (a SQLite performance fix), so the bugfix train still runs under the maintenance declaration. The README’s warning is the status fact that matters: the project is largely in maintenance mode, it will not accept new PRs or implement new features, and it names the dramatic change in frontier-model capabilities and a diversified research portfolio as the cause. The launch thread drew 282 points in July 2024, the largest graph-RAG footprint on Hacker News.
Strengths #
- Global sense-making is its product: community summaries answer “what are the themes across this corpus” questions that similarity search structurally cannot.
- It is the citation root of the pattern: LightRAG, nano-graphrag, and most graph-RAG work name it.
- MIT licensed with a CLI, a Python API, and docs deep enough to replicate the pipeline.
Cautions #
- Maintenance mode means the feature freeze is explicit: budget to own whatever you build on it, because upstream has said it is done.
- Indexing is expensive by the project’s own warning, an LLM call set per chunk over your whole corpus.
- Version bumps carry breaking config changes, and the README tells you to re-init between minor versions.
- Nothing agent-facing: no MCP server, no retrieval tools for a coding harness, the consumer is your own application.
Pricing #
Free and open source under MIT. The costs are your own LLM bills for indexing and querying, which the project’s documentation warns can be large on big corpora, so no prices belong in this note and no price history table applies.
Compared to #
- LightRAG: the actively maintained graph-RAG counterweight, dual-level retrieval instead of community summaries; pick it when you want graph RAG with a living release train.
- RAGFlow: the deployable engine with citations and a UI; graph structure is one artifact kind among its outputs, not the core.
- LlamaIndex: property-graph indexes give you the graph option inside a general framework, at the cost of assembling the pipeline yourself.
Bottom line #
Recommended for studying what graph RAG is and for one-shot theme extraction over a stable narrative corpus you can afford to index. Not for code, not for fast-changing corpora, and not as a dependency bet now that the maker has declared maintenance mode. My disagreeable claim: the maintenance-mode declaration is graph RAG’s candid status report, the frontier ate the gap it exploited, and the pattern’s heirs now argue about which parts survive as plain long-context questions.
Changes #
- 2026-10-07 - Created in the daily refresh’s retrieval entrant scan from the awesome-rag-production source.
- 2026-10-09 - Recorded PyPI 3.3.0 (2026-10-08, a SQLite performance fix, 51 releases) and refreshed the repository numbers (36,271 stars, pushed 2026-10-08); the README’s maintenance-mode declaration is unchanged.
- 2026-10-11 - Reworded a banned-term word out of a See also line and refreshed the repository numbers (36,288 stars); the maintenance-mode declaration is unchanged on re-read.
- 2026-10-22 - Rotation re-verification re-fetched every cited URL live (the maintenance-mode warning, the method paper, the docs, the launch thread all unchanged) and refreshed the repository activity (pushed 2026-10-20, 56 open issues, 52 PyPI releases, latest still 3.3.0).
- 2026-10-26 - Rotation re-verification re-fetched every cited URL live (the maintenance-mode warning, the method paper, the announcement post, the docs, and the launch thread all unchanged) and refreshed the repository activity (pushed 2026-10-26, 57 open issues); PyPI 3.3.0 remains latest.
- 2026-10-31 - Refreshed the repository activity (pushed 2026-10-29, 59 open issues); PyPI 3.3.0 and the maintenance-mode declaration unchanged.
- 2026-11-03 - Refreshed the repository activity (pushed 2026-11-02, 61 open issues; the default branch’s latest commit is a maintenance-workflow action-pinning change, inside the declared fix-only scope); PyPI 3.3.0 and the maintenance-mode declaration unchanged.
- 2026-11-07 - Rotation re-verification re-fetched every cited URL live (the repository API reading the same pushed date and 61 open issues, releases/latest still v3.3.0 of 2026-10-08, the README’s maintenance-mode and expensive-indexing warnings word-for-word, PyPI 3.3.0 across 52 releases, the method paper, the Microsoft Research announcement, the docs root, and the Algolia query still returning the cited 282-point launch thread); all claims unchanged.
- 2026-11-12 - Refreshed the repository activity (pushed 2026-11-09 at the repository level with the default branch’s latest commit still the 2026-11-02 workflow-action pinning, 64 open issues); releases/latest still v3.3.0 and the maintenance-mode declaration unchanged.
- 2026-11-15 - Rotation re-verification re-fetched every cited URL live (the repository API, releases/latest still v3.3.0 of 2026-10-08, the README’s maintenance-mode and expensive-indexing warnings and init-between-versions rule word-for-word, the method paper, the Microsoft Research announcement, the docs root, PyPI 3.3.0 across 52 releases, and the Algolia query still returning the cited 282-point launch thread); all claims unchanged, with the open-issue count refreshed from 64 to 67.
See also #
- LightRAG - the maintained graph-RAG implementation this project’s pattern spawned
- RAGFlow - the full RAG engine that packages graph-structured artifacts as one output among several
- Graphify - the context-engine sibling building agent-usable graph structure
- The Importance of Context When Interacting with LLMs - why retrieval structure beats raw model choice
References #
https://api.github.com/repos/microsoft/graphrag - 36,325 stars (as of 2026-10-13), 3,858 forks, MIT, created 2024-03-27, pushed 2026-11-09 at the repository level, the default branch’s latest commit dated 2026-11-02, 67 open issues (GitHub API, as of 2026-11-15)
https://api.github.com/repos/microsoft/graphrag/releases/latest - v3.3.0 published 2026-10-08, the SQLite performance-fix release (still latest, re-verified 2026-11-15)
https://raw.githubusercontent.com/microsoft/graphrag/main/README.md - the maintenance-mode warning, the expensive-indexing warning, the init-between-versions rule (re-verified unchanged 2026-11-15)
https://arxiv.org/abs/2404.16130 - the method paper: From Local to Global, graph RAG as query-focused summarization (re-verified live 2026-11-15)
https://www.microsoft.com/en-us/research/blog/graphrag-unlocking-llm-discovery-on-narrative-private-data/ - the Microsoft Research announcement grounding community summaries and the local/global query split (re-fetched 2026-11-15)
https://microsoft.github.io/graphrag/ - official docs root: local and global search, CLI quickstart (re-fetched 2026-11-15)
https://pypi.org/pypi/graphrag/json - version 3.3.0 (2026-10-08), 52 releases, MIT (re-verified 2026-11-15)
https://hn.algolia.com/api/v1/search?query=GraphRAG%20is%20now%20on%20GitHub&tags=story - the July 2024 launch thread, 282 points, 49 comments (re-fetched 2026-11-15, still the top hit)