GH Repository · gusye1234
nano-graphrag
A simple, easy-to-hack GraphRAG implementation
- stars
- 3,987
- 30-day movement
- +31/day
- Related entries
- 60
- Connections
- 1
Pythonllmpythongptlearning-by-doingraggraphraggpt-4o
nano-graphrag is a Python package implementing GraphRAG, a retrieval-augmented generation approach that uses graphs. It is described by its author as simple and easy to hack, with a learning-by-doing orientation.
You want a small, readable GraphRAG codebase you can study and modify rather than a heavy framework.
Use it to
- Study how GraphRAG works by reading the code
- Modify the implementation for your own experiments
- Build RAG pipelines with graph-based retrieval
- Prototype GraphRAG ideas with GPT models
For Developers learning or experimenting with GraphRAG
- Role
- rag
- Language
- Python
- Licence
- MIT
- Forks
- 426
- Open issues
- 68
- Last push
- 2026-01-27
- Latest release
- v0.0.5 · 2024-08-28
topicsgraphragragllmgptpythonlearning