A PDF to flashcards tool
that never sends the PDF anywhere.
Drop the file, the browser reads it, and every card keeps the sentence it came from.
No account, two decks a day, files up to 15MB and 25,000 characters free.
Open any question to see the line it came from. Paste 40 characters or more to start.
Test it before you paste anything
Open the PDF and try to click-drag select a line of it, the way you would select a sentence in an email. If a highlight follows your cursor, there is a text layer underneath the pixels and this tool will read it cleanly. If nothing highlights no matter where you click, you are looking at a picture of text rather than text, and no PDF-to-flashcards tool, ours included, can extract words that were never encoded as words in the first place.
Usually has a text layer: a PDF exported from Word, Google Docs or PowerPoint, most textbook PDFs bought digitally, lecture slides exported straight from Keynote, and anything you could already Ctrl+F inside your PDF reader.
Usually does not: a photocopied chapter run through a scanner, a photo of a whiteboard or a printed page saved as a PDF, and old library scans where every page is one large image. Those need OCR software first; this tool is not that, and pretending otherwise would just hand you an empty result with no explanation.
A 40-page PDF, start to finish
Drop it in
Use the upload button above the paste box, or drag the file onto it. A 40-page chapter is usually a few hundred kilobytes, well under the 15MB cap, and the extraction happens in a second or two, entirely in the tab you are looking at.
Watch what lands in the box
The extracted text fills the paste box the way it would if you had copied it by hand. If the chapter runs past 25,000 characters, the box tells you it trimmed to the first chunk rather than silently dropping the rest, so you know whether you are working from the whole thing or just the opening section.
Generate, then read before you export
Press generate and cards stream in, each with the line of the PDF it was pulled from underneath. A 40-page chapter usually gives you more candidate cards than are worth keeping; read through them once, and export the tab-separated file for Anki or a CSV once the set looks right.
The PDF itself never leaves your browser
Extraction runs client-side, using the same PDF rendering engine behind Firefox's built-in viewer, loaded into the page you are on. The file is read into memory, its text layer is pulled out page by page, and the PDF is discarded once that text lands in the paste box. Only the text goes anywhere from there.
You do not have to take our word for it. Open your browser's devtools, switch to the network panel, and drop a PDF onto the page: no upload request appears. For a lecture PDF with professor's names, unpublished data or anything else you would rather not hand to a server, that is the whole pitch.
Where this stops
No OCR, so a scanned or photographed PDF returns the "no text found" message instead of a guess. The free tier reads the first 25,000 characters of whatever you drop in; the rest is trimmed with a visible note, and reading a whole long PDF is a Pro waitlist feature, not something available today. The export is a tab-separated .txt or a CSV, not a packaged .apkg, and there is no image occlusion for the diagrams and figures a PDF often carries. What is left, extracting the words accurately and showing you where each card came from, is the part we tried to get right.
Before you drop the file in
Why did my scanned PDF come back empty?
A scanned page is a picture of text, not text. This tool reads the text layer a PDF carries when it was printed or exported from software; a photographed or scanned page has no such layer, so there is nothing to extract. We do not run OCR, so we say so plainly instead of returning something wrong.
Does the PDF get uploaded anywhere?
No. The file is opened and parsed by your browser, not sent to a server. Open your browser devtools, watch the network panel, and drop a PDF in: nothing fires. Only the plain text that ends up in the paste box travels anywhere, and that is the same request any pasted text would make.
What happens if the PDF is longer than the free limit?
The free tier reads up to 25,000 characters, roughly a chapter. A longer PDF gets trimmed to that many characters and the page tells you it did, so you are never guessing whether you got the whole thing. The full-length version is on the Pro waitlist.
Is there a file size limit?
Yes, 15MB. That is generous for a chapter or a lecture handout and tight for a scanned 300-page book, which would fail the text-layer check before size became the problem anyway.
What if the file won't open, or the tool says it's corrupt?
Some PDFs are password protected or damaged enough that the browser cannot parse them at all. The tool tells you that directly rather than failing silently. Open the file in a PDF reader first to confirm it opens normally, or copy the text out and paste it in instead.
How do I know a card didn't misread the PDF?
Every card shows the sentence it was built from, pulled from the extracted text. Open it and the source is right there next to the question, so a card that garbled a PDF footnote or a table is obvious before it ever reaches your deck.
Other ways in
Or start from the QuizPaste home page, where the same tool takes any text you paste.
QuizPaste is not affiliated with Adobe, Anki or AnkiWeb. PDF is a format, not a trademark we are claiming any relationship to; Anki is named only to describe the file format our export is compatible with.