Private, Modular
RAG Application
A comprehensive, Flutter-based Retrieval-Augmented Generation application built to be open-source, modular, and easy to deploy. Full control over your data and infrastructure.
Why Choose RAG.WTF?
We named it RAG.WTF because we wanted to answer "WTF is the RAG?" with unprecedented simplicity—and make setup so simple you'll say "Wow, That's Fast!"
Client-Side RAG
Runs entirely in the browser using SurrealDB WASM for a secure, serverless-optional experience.
Modular Monorepo
Built with Melos, separating concerns into distinct packages for better maintainability.
Multi-Provider LLM Support
Pre-configured for Ollama, OpenAI, Anthropic, Gemini, and more LLM providers.
Enhanced Data Privacy
User data can remain on the client machine, never being sent to a server.
Cost-Effective
Reduces server-side infrastructure needs, making it economical for personal use.
Cross-Platform
Built with Flutter for seamless deployment across web, mobile, and desktop.
Architectural Overview
The application follows a standard RAG pipeline, orchestrated across its modular packages for maximum flexibility and maintainability.
Document Ingestion
Upload documents through the UI, send to text-splitting service, chunk, and generate vector embeddings.
Storage & Indexing
Text chunks and vector embeddings are stored locally or remotely in SurrealDB for fast retrieval.
Retrieval & Generation
Query vectorization, similarity search, and context-aware answer generation using your chosen LLM.
Get Started in Minutes
Choose your setup based on your needs. From web application to full local deployment.
Web Application
Production-ready Single-Page Application
Local Setup
Maximum privacy & control
Join the Community
Connect with developers, share your projects, and get help from the RAG.WTF community.
Documentation
Comprehensive guides, tutorials, and API reference - Coming Soon!