# LightRAG

> A RAG framework with a knowledge graph and embeddings

- Category: Agents & dev (`agents`)
- Tags: `open-source`, `free`
- Official: https://github.com/HKUDS/LightRAG
- Wunderlust: https://wunderlust.xyz/r/lightrag
- Image: https://wunderlust.xyz/media/lightrag-poster.jpg
- Video preview: https://wunderlust.xyz/media/lightrag.webm
- Added: 2026-11-05

## What it is

An open source graph-based RAG framework that combines a knowledge graph with vectors, with a REST server and web interface.

## What it’s for

Use it when plain vector search isn't enough and you need to retrieve relationships between entities, for example in agent memory.

## Benefits

- Graph indexing with five query modes: local, global, hybrid, naive and mix
- Incremental updates and document deletion, with support for PDFs, images and tables
- Backends such as PostgreSQL, Neo4j, Milvus and Qdrant

## Caveats

- MIT license; the costs of the connected LLM and embeddings are not covered by the project
- The default in-memory stores are for testing only, according to the repository
- The video says LightRAG; the repository is only possibly the project mentioned

## Similar on Wunderlust

- [Supabase](https://wunderlust.xyz/r/supabase.md): Open-source backend platform built on Postgres
- [Unsloth Studio](https://wunderlust.xyz/r/unsloth-studio.md): App for running, chatting with and training AI models locally
- [Gbrain](https://wunderlust.xyz/r/gbrain.md): Second-brain tool for AI agents
- [OpenViking](https://wunderlust.xyz/r/openviking.md): Context database for agents that organizes memory and skills like a file system
- [Jev Ultrafast](https://wunderlust.xyz/r/jev-ultrafast.md): A Python browser agent that picks indexed actions over the DOM
- [Agent Browser](https://wunderlust.xyz/r/agent-browser.md): Tool that gives AI agents web browsing