Summarize any PDF in about ten seconds.
Drop the file in. It is read in your browser, never uploaded, and you get back the points that actually carry the document.
How it works
- Drop the file. Your browser extracts the text with pdf.js. The PDF itself never touches a server, ours or anyone else's.
- Only the text travels. We send the extracted words to a language model, get the points back, and keep nothing.
- Read the brief. Five points by default, ten if you want the detail.
Why the file stays on your device
Most PDF summarizers ask you to upload. That is the easy way to build one, and it means your document sits on someone else's disk with whatever retention policy they happen to have. Contracts, medical records, internal decks, unpublished drafts: none of that should be casually handed to a third party.
Extraction in the browser removes the question entirely. The only thing that leaves your machine is plain text, and only when you press the button.
What this handles well, and what it does not
Works well on: reports, whitepapers, contracts, book chapters, slide decks exported to PDF, anything with a selectable text layer.
Struggles with:
- Scanned documents. A photo of a page has no text layer. Nothing to extract, so nothing to summarize. You would need OCR first.
- Heavy tables and figures. The numbers come through as a jumble of loose cells, and a summary of a jumble is not worth much.
- Very long documents. We read roughly the first 48,000 characters, which is around 60 to 80 pages of normal prose. A 400-page book gets its opening summarized, not its argument.
If a file comes back with "no text in this PDF", it is almost always the scanned-image case.
Questions
Is my PDF uploaded anywhere?
No. The text is extracted by your own browser using pdf.js, and only that extracted text is sent for summarizing. The file itself never leaves your device.
How many pages can it handle?
Around 60 to 80 pages of normal prose per brief. Beyond that we summarize the first portion and tell you how many pages were used.
Why did it say there is no text in my PDF?
The document is almost certainly a scan, meaning each page is an image rather than text. Run it through OCR first, then try again.