RAG Web UI provides a retrieval-augmented generation based dialogue system with a TypeScript frontend and Python/FastAPI backend. It supports multiple LLM providers, embedding options, and vector stores, with Docker-based quick start.
Collecting history — the radar snapshots this repo daily. The trend line appears after 3 days of data (1 so far).
What it is
RAG Web UI is a knowledge base management system based on Retrieval-Augmented Generation (RAG) technology. The project describes a system that combines document retrieval and large language models to enable knowledge-based question answering with OpenAI, DeepSeek, MiniMax, Ollama, and local embedding options, exposed via an API.
How it works
The README shows a frontend-backend architecture that includes FastAPI, MinIO, MySQL, ChromaDB/Qdrant as vector databases, LangChain, and a frontend built with Next.js 14. It describes an embedding service and a retrieval workflow involving document ingestion, vector storage, cross-encoder reranking, context assembly, and LLM generation. The architecture diagram and sections imply a pipeline from document ingestion to user-facing responses with citations.
Getting started
Prerequisites: Docker & Docker Compose v2.0+, Node.js 18+, Python 3.9+, 8GB+ RAM. Installation steps include:
- git clone https://github.com/rag-web-ui/rag-web-ui.git
- cd rag-web-ui
- cp .env.example .env
- docker compose up -d --build Development guide suggests:
- docker compose -f docker-compose.dev.yml up -d --build Configuration is split into Core, LLM, Embedding, Vector Database, and Object Storage sections with various parameters and defaults shown; sample values include MYSQL_SERVER=localhost, SECRET_KEY, OPENAI_API_BASE, etc.
Recent releases
Latest release: v0.8.0 (2026-04-06) adding HuggingFace Embeddings Support (EMBEDDINGS_PROVIDER=huggingface) via sentence-transformers. Other listed releases include v0.7.5, v0.7.3, v0.7.2, v0.7.1 with fixes and enhancements such as bcrypt integration and Docker host configuration.
Traction
Stars: 3087.
Behind the repo
Not provided in the README FACTS block beyond repository owner and structure; no startup/company link is specified.
Caveats
License: Apache-2.0. Created: 2025-01-14. Last push: 2026-04-06. Open issues: 17. Language: TypeScript. Project notes indicate it is for learning/sharing RAG knowledge and not production-ready. For deployment, image relies on Docker and multiple services (MySQL, MinIO, vector DBs).