PDF.ai
PDF.ai lets you chat with any PDF. You upload a document, ask questions in plain language and get answers, summaries and extracted details, each linked to the page it came from. The same company also sells a document processing API that turns PDFs into structured data for developers.
It suits students, researchers, lawyers, finance teams and anyone who reads long documents and wants to find the relevant part quickly. Developers who need to pull text, tables or specific fields out of PDFs at scale are the audience for the API.

PDF.ai's public demo: the document on the left, the answer with page citations on the right.
Chat with your PDFs
The app shows your PDF beside a chat panel. Ask for a summary, a definition, a clause or a figure, and the answer comes back with citation buttons such as "p. 1" that jump to the source page. Answers come from the document you uploaded rather than the open web, which makes them easy to check. Higher plans let you chat across all your PDFs at once, transcribe voice questions and read answers aloud, and use OCR to handle scanned pages.
PDF.ai is also available as a Chrome extension, so you can open a PDF in the browser and chat with it without uploading it to the main app first.
AI document agents
Paid plans from Ultimate up include specialized document agents for legal, finance, real estate, healthcare, education and research work. Each one is set up to analyze that type of document, such as contracts or financial reports, and answer questions about it.
PDF API for developers
The API has three main endpoints. Parse uses OCR and layout detection to turn a PDF into structured output with headings, paragraphs, tables and figures. Extract pulls specific fields, such as names, dates or amounts, using your own prompts. Split divides a document into the sections you define. There's also a chat endpoint for asking questions about uploaded documents. PDF.ai says most documents return in under five seconds and that files are deleted after processing.
Free PDF tools
The site has a set of free tools that work without an account, including merge, compress, split, flatten, page numbering and conversions between PDF, Word, Markdown, PNG and JPEG.
Plans and pricing
The chat app has a free Hobby plan with one upload and a limited number of questions each month. Pro raises those limits a long way, while Ultimate adds unlimited uploads and questions, team invites, the AI agents and better audio models. Enterprise adds larger files and a white-labeled PDF embed. Paying yearly costs less than paying monthly.
The API is priced separately on monthly credit plans, with a free tier for testing and higher tiers for production volume. Custom plans are available above the largest tier.
Is PDF.ai free?
Yes, with limits. The Hobby plan is free and needs no credit card, but it allows only one upload a month. The API also has a free monthly credit allowance for testing.
Does PDF.ai show where answers come from?
Yes. Each answer includes citation buttons that point to the page in your document where the information appears.
Can PDF.ai read scanned PDFs?
Yes. It uses OCR on scanned pages, with a page limit per file that rises on higher plans.
Is there a PDF.ai API?
Yes. The API can parse PDFs into structured data, extract custom fields, split documents into sections and answer questions about uploaded files. It has its own credit-based plans, separate from the chat app.
Alternatives to PDF.ai
- Foxit: Full PDF editor with AI document chat, summaries, translation and multi-document research with source citations.
- Anara: AI research workspace that reads PDFs, audio and video, then writes with clickable citations across many sources.
- NoteGPT: Turns PDFs, videos and slides into summaries, notes, flashcards and mind maps, with AI chat.
- Thunderbit: Extracts structured tables from websites, PDFs and images with AI field detection.

TRY IT