openrag

OpenRAG is a comprehensive, single-package Retrieval-Augmented Generation (RAG) platform for intelligent document search and AI-powered conversations. It is built with FastAPI and Next.js and powered by OpenSearch, Langflow and Docling. Key capabilities described in the README: - Pre-packaged and ready to run: core tools are wired together so users can install and start quickly. - Agentic RAG workflows: orchestration with re-ranking and multi-agent coordination. - Document ingestion: parsing of messy, real-world data via Docling. - Drag-and-drop workflow builder: visual interface powered by Langflow for rapid iteration. - Modular enterprise add-ons and enterprise search at scale via OpenSearch. Typical workflow: launch OpenRAG, add files or folders as knowledge, then chat with that knowledge through a chat interface backed by LLMs and semantic search. Integration options include official Python and TypeScript/JavaScript SDKs (openrag-sdk) for programmatic chat and search, plus a built-in Model Context Protocol (MCP) server over streamable HTTP mounted at /mcp. MCP clients such as Cursor and Claude Desktop can connect using the OpenRAG API key via the X-API-Key header; the README notes the standalone openrag-mcp PyPI package is deprecated in favor of the /mcp endpoint. The MCP server exposes tools for RAG-enhanced chat, semantic search, document ingestion, knowledge filters and settings management. Installation paths documented include a quickstart, installing the OpenRAG Python package, and deploying self-managed services with Docker or Podman. Development and contribution guidance is in CONTRIBUTING.md, with troubleshooting and discussions available through the project documentation and GitHub pages.