# HKUDS/LightRAG

[EMNLP2025] LightRAG: Simple and Fast Retrieval-Augmented Generation

Repository: https://github.com/HKUDS/LightRAG
Canonical: https://ross.abutalabs.com/products/lightrag
Homepage: https://arxiv.org/abs/2410.05779
Language: Python
License: MIT
License Family: permissive
Topics: knowledge-graph, large-language-models, retrieval-augmented-generation, genai, graphrag, llm, rag, gpt, docling, mineru, ragas
Last push: 2026-08-26T00:51:09+00:00

## Health v2 (maintenance only)
Score: 85/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 99, release rhythm 86, longevity 50
- inputs: {"age_days": 700, "days_push": 8, "days_rel": 14, "gap_med": 5.0, "n_releases_24m": 79}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 39204, forks 5510 (observed 2026-08-28T04:12:07.775030+00:00)

## What it is
LightRAG is a Python framework for retrieval-augmented generation that integrates graph structures into text indexing and retrieval, using a dual-level retrieval system combining knowledge graphs with vector representations. It supports incremental data updates and offers improved retrieval accuracy and efficiency over flat-representation RAG systems.

## Use cases
- build a RAG pipeline over my documents
- question answering over a knowledge base with graph-based retrieval
- index large document collections for LLM retrieval
- graph-based RAG instead of flat vector search
- incrementally update a RAG index as new documents arrive
- chat with my PDFs using a knowledge graph

## When to choose
- you need RAG that captures entity relationships and complex inter-dependencies between documents
- you want fast, graph-enhanced retrieval with incremental index updates
- you want a simple, well-adopted Python RAG library with an active community

## When to avoid
- you need only simple flat vector similarity search without graph overhead
- you want a fully managed hosted RAG service rather than a self-run library
- your project is not Python-based

## Facets
- artifact type: library
- maturity: active
- function: rag, search-engine, llm-inference, nlp
- domain: large-language-models, artificial-intelligence
- platform: python, cross-platform
- tags: knowledge-graph, graphrag, graph-based-retrieval, vector-search, document-indexing, llm, retrieval-augmented-generation, natural-language-processing

## Member repositories
- HKUDS/LightRAG (main) score 85

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:12:07.775030+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T16:24:17.188962+00:00, confidence not recorded.
  - readme: https://github.com/HKUDS/LightRAG (fetched 2026-08-28T04:12:07.775030+00:00, sha 78421ae22387)
  - homepage: https://arxiv.org/abs/2410.05779 (fetched 2026-08-29T07:46:53.197845+00:00, sha 24f607bee105)
  - site_page: https://info.arxiv.org/about/donate.html (fetched 2026-08-29T07:46:53.207210+00:00, sha cca9c3a11c56)
  - site_page: https://info.arxiv.org/about/ourmembers.html (fetched 2026-08-29T07:46:53.210954+00:00, sha 47cbc55ff1de)
  - site_page: https://info.arxiv.org/about (fetched 2026-08-29T07:46:53.212684+00:00, sha a1f16f915a9a)
  - site_page: https://info.arxiv.org/labs/index.html (fetched 2026-08-29T07:46:53.209174+00:00, sha b14a8d05a0ec)
- Data as of 2026-08-30T08:39:29.467469+00:00.