RAGFlow
Deep-document-understanding RAG engine from a Chinese team, for enterprise document Q&A
A RAG engine with deep document understanding from a Chinese team, built for enterprise-grade document Q&A.
Frameworks and platforms for building autonomous AI agents and LLM apps.
Key features
- Open-source RAG engine fusing RAG with agent capabilities
- Deep document understanding: layout analysis, OCR, tables
- Template-based chunking and agentic multi-step retrieval
- Compiles knowledge into wikis, graphs, mind maps, timelines
- Go-native services with optional MCP and sandbox executor
Good for
- Enterprise RAG over complex unstructured documents
- Developers needing grounded citations and traceability
Stack & tags
Goragdocument-understandingknowledge-baseagentic-retrievalocrcitationsself-hostedgo
Screenshots
Screenshots from the project’s GitHub README.
Similar open-source projects
superpowers
★ 294,114An agentic skills framework & software development methodology that works.
skills
★ 274,171Skills for Real Engineers. Straight from my .agents directory.