RAG Web UI is an intelligent dialogue system based on RAG (Retrieval-Augmented Generation) technology that helps build intelligent Q&A systems based on your own knowledge base. By combining document retrieval and large language models, it achieves accurate and reliable knowledge-based question answering services.

The system supports multiple LLM deployment options, including cloud services like OpenAI and DeepSeek, as well as local model deployment through Ollama, meeting privacy and cost requirements in different scenarios.

It also provides OpenAPI interfaces for convenient knowledge base access via API calls.

• 📚 Intelligent Document Management

  • Support for multiple document formats (PDF, DOCX, Markdown, Text)
  • Automatic document chunking and vectorization
  • Support for async document processing and incremental updates

• 🤖 Advanced Dialogue Engine

  • Precise retrieval and generation based on RAG
  • Support for multi-turn contextual dialogue
  • Support for reference citations in conversations

• 🎯 Robust Architecture

  • Frontend-backend separation design
  • Distributed file storage
  • High-performance vector database: Support for ChromaDB, Qdrant with easy switching through Factory pattern

• Docker & Docker Compose v2.0+
• Node.js 18+
• Python 3.9+
• 8GB+ RAM

1. Clone the repository

git clone https://github.com/rag-web-ui/rag-web-ui.git
cd rag-web-ui

2. Configure environment variables

You can check the details in the configuration table below.

cp .env.example .env

3. Start services(development server)

docker compose up -d --build

Access the following URLs after service startup:

• 🌐 Frontend UI: http://127.0.0.1.nip.io
• 📚 API Documentation: http://127.0.0.1.nip.io/redoc
• 💾 MinIO Console: http://127.0.0.1.nip.io:9001

• 🐍 Python FastAPI: High-performance async web framework
• 🗄️ MySQL + ChromaDB: Relational + Vector databases
• 📦 MinIO: Distributed object storage
• 🔗 Langchain: LLM application framework
• 🔒 JWT + OAuth2: Authentication

• ⚛️ Next.js 14: React framework
• 📘 TypeScript: Type safety
• 🎨 Tailwind CSS: Utility-first CSS
• 🎯 Shadcn/UI: High-quality components
• 🤖 Vercel AI SDK: AI integration

The system is optimized in the following aspects:

• ⚡️ Incremental document processing and async chunking
• 🔄 Streaming responses and real-time feedback
• 📑 Vector database performance tuning
• 🎯 Distributed task processing

docker compose -f docker-compose.dev.yml up -d --build

Parameter	Description	Default	Required
MYSQL_SERVER	MySQL Server Address	localhost	✅
MYSQL_USER	MySQL Username	postgres	✅
MYSQL_PASSWORD	MySQL Password	postgres	✅
MYSQL_DATABASE	MySQL Database Name	ragwebui	✅
SECRET_KEY	JWT Secret Key	-	✅
ACCESS_TOKEN_EXPIRE_MINUTES	JWT Token Expiry (minutes)	30	✅

Parameter	Description	Default	Applicable
CHAT_PROVIDER	LLM Service Provider	openai	✅
OPENAI_API_KEY	OpenAI API Key	-	Required for OpenAI
OPENAI_API_BASE	OpenAI API Base URL	https://api.openai.com/v1	Optional for OpenAI
OPENAI_MODEL	OpenAI Model Name	gpt-4	Required for OpenAI
DEEPSEEK_API_KEY	DeepSeek API Key	-	Required for DeepSeek
DEEPSEEK_API_BASE	DeepSeek API Base URL	-	Required for DeepSeek
DEEPSEEK_MODEL	DeepSeek Model Name	-	Required for DeepSeek
OLLAMA_API_BASE	Ollama API Base URL	http://localhost:11434	Required for Ollama
OLLAMA_MODEL	Ollama Model Name	llama2	Required for Ollama

Parameter	Description	Default	Applicable
EMBEDDINGS_PROVIDER	Embedding Service Provider	openai	✅
OPENAI_API_KEY	OpenAI API Key	-	Required for OpenAI Embedding
OPENAI_EMBEDDINGS_MODEL	OpenAI Embedding Model	text-embedding-ada-002	Required for OpenAI Embedding
DASH_SCOPE_API_KEY	DashScope API Key	-	Required for DashScope
DASH_SCOPE_EMBEDDINGS_MODEL	DashScope Embedding Model	-	Required for DashScope
OLLAMA_EMBEDDINGS_MODEL	Ollama Embedding Model	deepseek-r1:7b	Required for Ollama Embedding

Parameter	Description	Default	Applicable
VECTOR_STORE_TYPE	Vector Store Type	chroma	✅
CHROMA_DB_HOST	ChromaDB Server Address	localhost	Required for ChromaDB
CHROMA_DB_PORT	ChromaDB Port	8000	Required for ChromaDB
QDRANT_URL	Qdrant Vector Store URL	http://localhost:6333	Required for Qdrant
QDRANT_PREFER_GRPC	Prefer gRPC Connection for Qdrant	true	Optional for Qdrant

Parameter	Description	Default	Required
MINIO_ENDPOINT	MinIO Server Address	localhost:9000	✅
MINIO_ACCESS_KEY	MinIO Access Key	minioadmin	✅
MINIO_SECRET_KEY	MinIO Secret Key	minioadmin	✅
MINIO_BUCKET_NAME	MinIO Bucket Name	documents	✅

Parameter	Description	Default	Required
TZ	Timezone Setting	Asia/Shanghai	❌

We welcome community contributions!

1. Fork the repository
2. Create a feature branch (git checkout -b feature/AmazingFeature)
3. Commit changes (git commit -m 'Add some AmazingFeature')
4. Push to branch (git push origin feature/AmazingFeature)
5. Create a Pull Request

• Follow Python PEP 8 coding standards
• Follow Conventional Commits

• Knowledge Base API Integration
• Workflow By Natural Language
• Multi-path Retrieval
• Support Multiple Models
• Support Multiple Vector Databases

This project is licensed under the Apache-2.0 License

This project is for learning and sharing RAG knowledge only. Please do not use it for commercial purposes. It is not ready for production use and is still under active development.

Thanks to these open source projects:

• FastAPI
• Langchain
• Next.js
• ChromaDB

---