Chat with PDF
Upload a document and ask anything — summaries, key figures, specific clauses — with answers that cite the page. Your file stays in your browser; only the question and extracted text go to your AI provider.
Your file never leaves your deviceDrop a PDF to chat about
or click to browse — text-based PDFs only (for scans, OCR them first)
Azure OpenAI chat-completions URL, or any OpenAI-compatible endpoint (OpenRouter, Groq, local LLM…).
The key is stored only in this browser (localStorage) and sent directly to your provider. For Azure, enable CORS for this origin in the Azure portal (Resource Management → CORS).
Chat with your document
Ask anything.
Try: "Summarize this document in 3 bullets" or "What are the key dates mentioned?"
Try: "Summarize this document in 3 bullets" or "What are the key dates mentioned?"
💡 How it works: the document text is extracted locally with page markers, then sent with your question to the AI provider you configure. Answers reference pages like (p. 5). No document is ever stored or uploaded to PDFLift.