LightRAG on Ubuntu 24.04

Overview

LightRAG is an open source retrieval augmented generation server from the HKU Data Intelligence Lab. You upload your own documents - PDFs, Word files, PowerPoint decks, spreadsheets, markdown or plain text - and it uses your chosen language model to pull the entities and the relationships between them into a knowledge graph, while embedding the same text into a vector index. Questions are then answered from both at once, so an answer is grounded in how the facts in your corpus connect rather than in whichever paragraphs merely looked similar. The web interface lets you upload and track documents, watch the indexing pipeline, explore the extracted graph visually, edit entities and relationships by hand, and query in naive, local, global, hybrid or mix modes; everything is also available over a documented REST API and an Ollama compatible chat endpoint. LightRAG is a front end to your models rather than a model host, so it runs comfortably on a small CPU only VM with no GPU: you point it at OpenAI, Azure OpenAI, Gemini, Bedrock, Anthropic or a private Ollama, vLLM or LM Studio endpoint and keep control of the model and what it costs.

Common uses

• Ask grounded questions across a private research or documentation corpus
• Build a knowledge graph from contracts, policies or technical manuals
• Run a self hosted RAG backend behind your own model provider account

See it running

Real screenshots taken while testing this image against its deployment guide.