Chat with Documents

​

Features

• PDF Document Upload and Preview
• Interactive Chat Interface
• Powered by Qwen3-235B-A22B Model
• Advanced RAG Implementation using LlamaIndex
• High-quality Embeddings with BAAI/bge-en-icl
• Real-time Document Processing
• Transparent AI Reasoning Display

​

Prerequisites

• Python 3.10
• Nebius AI API Keys

​

Setup

1. Clone the repository:
   Copy

   Ask AI

   git   clone    https://github.com/nebius/ai-studio-cookbook/
   cd rag/chat-with-pdf
   

2. Install the required packages: If using uv :
   Copy

   Ask AI

   uv sync
   

   If using pip:
   Copy

   Ask AI

   pip install -r requirements.txt
   

3. Set up your environment variables: Create a .env file in the project root and add your Nebius API key:
   Copy

   Ask AI

   NEBIUS_API_KEY=your_api_key_here
   

​

Running

1. Run the Streamlit application: If using uv :
   Copy

   Ask AI

   uv run streamlit  run main.py
   

   If using pip:
   Copy

   Ask AI

   streamlit run main.py
   

2. Open your web browser and navigate to the provided local URL (typically http://localhost:8501)
3. Upload a PDF document using the sidebar
4. Start chatting with your document!

​

Features in Detail

​

Document Processing

• Supports PDF file uploads
• Real-time document preview in the sidebar
• Automatic document indexing using LlamaIndex

​

Chat Interface

• Clean and intuitive chat UI
• Support for multiple message types
• Clear chat history functionality
• Expandable AI reasoning display

​

Model Options

• Primary: Qwen3-235B-A22B
• Alternative: DeepSeek-V3
• Embedding Model: BAAI/bge-en-icl

​

Architecture

• Streamlit for the web interface
• LlamaIndex for document processing and RAG implementation
• Nebius AI’s models for embeddings and generation
• PyPDF2 for PDF handling

​

Contributing

​

Credits