Ask questions from any PDF using OpenAI + LangChain + Chroma.
This project lets you upload a PDF and ask natural language questions about its content. It uses:
git clone https://github.com/yourname/langchain-pdf-qa.git
cd langchain-pdf-qacurl -Ls https://astral.sh/uv/install.sh | shuv syncCreate a .env file based on the example:
cp .env.example .envThen edit .env and replace with your API key.
uv run python ask_pdf.pyPlace a PDF file (e.g., example.pdf) in the root and run the script. You’ll be prompted to ask questions like:
- "What is this document about?"
- "Who is the author?"
- "Summarize the key findings."
Type exit to quit.
Ask a question (or type 'exit'): What is the main purpose of this document?
Answer: This document provides an overview of ...
.
├── ask_pdf.py
├── example.pdf # Your PDF file (not committed)
├── .env # Your API key (not committed)
├── .env.example
├── pyproject.toml
├── README.md
├── .gitignore
Do not commit your .env file or PDF files to GitHub. Use .gitignore to keep them private.
- Add a Streamlit or Flask UI
- Use FAISS or Pinecone as vector store
- Deploy to Hugging Face Spaces or Render
This project uses uv for dependency management and Python environment handling.
# Install dependencies
uv sync
# Run a script
uv run python ask_pdf.py
# Add a new dependency
uv pip install package-name
# Remove a dependency
uv pip uninstall package-name
# Update dependencies
uv sync --upgrade
# Run tests
uv run python -m pytest
# Activate the virtual environment (if needed)
source .venv/bin/activate- Fast: Much faster than pip and Poetry
- Reliable: Better dependency resolution
- Simple: Single tool for environment management, dependency installation, and running scripts
- Modern: Built for Python 3.13+ with modern tooling
MIT — free to use and modify.