PDF research tool. Drop in PDFs, they get chunked, embedded, and images extracted+described in the background. Then select text in the viewer and ask questions — answered by Claude Code searching the vector DB and surrounding context.
- OpenAI API key — for embeddings and image descriptions
- Claude Pro/Max subscription — for Q&A via Claude Code
- Claude Code — must be installed and authenticated (
npm install -g @anthropic-ai/claude-code)
cp .env.example .env
# add your OPENAI_API_KEY to .env
uv sync
uv run uvicorn server:app --reload --port 8000
Open http://localhost:8000, drop a PDF, wait for ingestion, select text, ask questions.