LightRAG
by HKUDS
LightRAG is a Python framework and server that lets an AI language model answer questions using your own documents. It builds a knowledge graph from your files and searches it, and it includes a web interface for adding documents, asking questions and viewing the results.
- Stars
- 40.5k
- Licence
- MIT
- Runs on
- A small server
- Setup
- Technical
- Language
- Python
- Age
- 760 days
What you can do with it
- Load PDFs, Office files and other documents and ask questions answered from their content, with source citations
- Explore the knowledge graph built from your documents using the web dashboard
- Connect it to a choice of language models and storage systems such as PostgreSQL, Neo4j, MongoDB or OpenSearch
Make money with it
A freelancer could set up and maintain private document question-and-answer systems for small firms as a fixed-fee or monthly service.
An idea, not financial advice. Check the licence before charging for anything built on it.
Ways a business could use it
- Build an internal assistant that answers staff questions from manuals, policies and past project documents
- Offer clients a searchable question-and-answer system over their own document collections as a paid service
Can you use it commercially?
Free to use, change and sell, including in paid products. Keep the copyright notice.
Not legal advice. Read the project's licence file before building a product on it.
Getting it running
Install the LightRAG server from PyPI with uv or pip (a setup wizard and Docker options for local components are also mentioned), then configure a language model and embedding service.
Follow the official instructions in the project README.
Things to check
- Needs a language model and embedding service, which usually means paid API keys unless you run models yourself
- Setup needs Python and some technical knowledge
- Indexing large document sets can cost a lot in model usage
Who's talking about it
Checked 02 Nov 2026. Summary drafted by AI from the project's README.
