1 · Source

Drop a document here or tap to choose a file

PDFDOCXXLSXXLSCSVPNGJPG

The free version reads up to about 10 pages of each document.

How to read PDFs

Text layerFast. Scanned pages are read with OCR automatically. Only what's visible on the pageStrictest. Every page is drawn and read with OCR, so hidden or white text is left behind.

Include headers, footers and footnotes

Watermark check

Metadata in the original

What was left behind

2 · Clean text

Edit freely. Lines starting with # print as headings.

3 · Rewrite (optional)

Say the same thing in different words

Changes wording and sentence structure so a copy can't be matched by phrasing. Facts, names, numbers and order stay the same.

Why rewrite? Creating the clean PDF already removes marks hidden in spacing, shading and characters. But a document can also be marked by its word choices, which look completely normal, and only new wording removes those. See how wording marks work

Rewritten by Claude. Your text is sent to Anthropic through this site to be rewritten; the original file is not. Claude's output carries Anthropic's statistical watermark, which shows Claude was involved but not who used it.

Writing style

4 · New PDF

Build a clean PDF

A brand-new file set in standard fonts. It can follow the original's headings, bold and italic, lists and tables, but no author, dates, embedded fonts, images, colors or exact spacing carry over.

Paper

Text size

Formatting Spreadsheets and text files are read without formatting, so this copy uses the plain layout.

Title on page 1 (optional)

File name

Page numbers Add a few human-style typos Never touches numbers, dates, names, addresses, headings or table rows. Different every time.

The pages of your new PDF will show here after you create it.

What Unmark PDF removes

How do I remove metadata from a PDF?

Open the file on this page. Unmark PDF lists the metadata it finds, such as the author, the software used, dates and edit history. Then it makes a new PDF that carries none of it, with blank properties and a fixed creation date.

Can it remove hidden text from a PDF?

Yes. Choose “Only what's visible on the page” and the PDF is read the way it looks, so white or hidden words are left behind. Without that option, hidden words come through as ordinary text, so you can spot and delete them.

What is an invisible watermark?

A small change you can't easily see, such as uneven word spacing, a faint dot pattern or hidden characters, that makes each copy of a document slightly different. Unmark PDF checks for signs of these marks, then sets the words again on a plain white page. No tool can promise a document can never be traced, so read the result before you share it.

Which tools add watermarks to documents?

Virtual data rooms, secure file sharing services, document security (DRM) software and PDF editors can all add them. Some put the viewer's name or email on each page, which is often called a dynamic watermark. Others hide a code in the spacing or shading. A mark you can see may come through as ordinary text, so read the result before you share it.

How do I clean a Word document before sharing it?

Open the .docx file here. You get a new PDF, not a Word file, without the author and company details, comments, deleted tracked changes, hidden text or revision IDs.

Does it work on Excel files and pictures?

Yes. From Excel, it leaves out hidden sheets, rows and columns, comments and formulas, and keeps the values you can see. Pictures are read with OCR, so only the words are kept and any camera or GPS data is gone.

Is my file uploaded?

No. Reading, OCR and PDF creation all run in your browser. Only the optional Claude rewrite sends text out: the text in the editor goes to Anthropic through this site, and Claude's output carries Anthropic's statistical watermark.

Is it free?

Yes, for documents up to about 10 pages. With a longer document, only the first part is read, and the report tells you what was left out.