[EMNLP2025] LightRAG: Simple and Fast Retrieval-Augmented Generation
LightRAG is a retrieval-augmented generation system that combines language models with retrieved information and knowledge-graph data to answer questions using relevant sources. It is for building applications that search document or knowledge collections before generating responses.
Latest release v1.5.7 · 2 Sept 2026
4 files for Claude Code, Codex and OpenCode: SessionStart, LightRAG, LightRAG AGENTS.md, LightRAG CLAUDE.md — 7,651 tokens loaded in every session.
▶ Claude Code + LightRAG = UNSTOPPABLE Chase AI · about HKUDS/LightRAG · on YouTube → AGENTS.md A 7,646 tok CLAUDE.md A 5 tok .claude/settings.json A — .claude/settings.json A — These files are HKUDS/LightRAG's own configuration — they tell Claude Code, Codex and OpenCode how to work on this repository, so they are not mods to install elsewhere. Copy one as a starting point and replace the parts that are about this project.