Vibe Coding DiscoverLightRAG
View on GitHub[EMNLP2025] LightRAG: Simple and Fast Retrieval-Augmented Generation
LightRAG is a Python RAG framework that builds a dual-level knowledge graph over documents for entity- and relation-aware retrieval. It ships an API server, WebUI, multiple storage backends (Neo4j, Postgres, Mongo, Milvus, OpenSearch), reranking, multimodal parsing and OpenAI/Ollama/Gemini integrations.
Use Cases
Graph-based document question answeringBuilding knowledge graphs from raw textEnterprise knowledge base retrievalMultimodal RAG over PDFs, images and tablesCitation-backed answer generationLocal/offline LLM RAG deploymentRAG evaluation with RAGAS and tracing with LangfuseCustom KG insertion for domain data
Built With
- Language
- Python
- Frameworks
- networkx · nano-vectordb · neo4j · postgresql · mongodb · milvus · opensearch · langfuse · ragas · docling · mineru · ollama · docker · uv
Tags
rag · knowledge-graph · graphrag · retrieval-augmented-generation · llm · embeddings · vector-database · hybrid-retrieval · reranking · multimodal · document-parsing · citations · self-hosted · webui · neo4j · chunking