About RAGFlow
RAGFlow is an open source retrieval augmented generation engine. It focuses on parsing documents carefully before an AI system searches them.
What it does
RAGFlow ingests files such as PDFs, images and spreadsheets and turns them into structured, searchable data. It combines vector search, full text search and re-ranking to find the most relevant passages. The platform also includes a visual builder for agents that use retrieval, tools and MCP servers in one workflow.
Key features
- Document ingestion for many file types
- Hybrid search with vector and full text
- Re-ranking for more accurate answers
- Visual builder for agent workflows
- Connects to MCP servers and tools
Who it's for
Engineering and knowledge teams that build question answering over internal documents. The site names enterprise teams with large document sets as its main audience.
Pricing and good to know
The code is licensed under Apache 2.0. Model and infrastructure costs depend on the models and hosting you choose. The site has a Pricing page and a contact option for enterprise needs, and it lists web search and chat among its features.
Highlights
- Parse PDFs, images and spreadsheets
- Search by vector and full text together
- Re-rank results for better answers
- Build agent workflows visually
- Call MCP servers and tools from a workflow
