Ask grounded questions about a PDF.
Upload a text-based PDF, review the extracted source status, and ask focused questions inside a document-first AI workspace.
Chat with PDF — FAQ
How does it work?
Upload a text-based PDF and DockDocs extracts its text in your browser, then you ask questions and the AI answers using that document's content — not generic web knowledge. It's a focused MVP: one document at a time, up to 12 pages / 40,000 characters / 25 MB per file.
Can I trust the answers — are they tied to my document?
Answers are grounded in the text extracted from your PDF, and when the AI's reply matches passages in your file, those appear under a '✓ verified against source' note with the exact quotes, so you can check them yourself. Citations depend on what the model returns and won't appear for every answer, and AI can still be wrong — for anything important, confirm against the original document.
Is my PDF uploaded or stored?
The text is extracted in your browser; only that extracted text is sent to the AI provider to answer your question, and the file itself never leaves your device and isn't stored afterwards.
Which PDFs work?
Text-based (born-digital) PDFs. Scanned PDFs have no selectable text — run OCR first so chat has something to read. It works in English, Chinese, Spanish, Portuguese, French and more; answers follow the language you ask in.
Are there limits?
The MVP caps each file at 25 MB, 12 pages, and 40,000 characters of extracted text; for longer documents, split or compress them first. A daily free quota applies — upgrade if you need more.
Chat with PDF questions
What is DockDocs Chat with PDF?+
DockDocs Chat with PDF lets users upload a PDF, extract readable text in the browser, and ask questions through a configured AI provider.
Does it answer without an AI provider?+
No. If the provider endpoint or API key is unavailable, the page returns a real error instead of a simulated answer.
Does it support scanned PDFs?+
This MVP reads selectable PDF text. Scanned PDFs require OCR before chat can answer against their content.
Is it static export compatible?+
Yes. The page is static-export compatible, while provider calls are routed through a Netlify Function when deployed.