How to redact a PDF
- Choose your PDF above, or tap Try a sample. It opens in your browser; nothing is uploaded.
- Type a name or number into Find and redact, tick the patterns you want (emails, phone numbers, SSNs, account numbers, dates) and tap Find matches.
- Drag on the page to black out anything else, such as a signature, a photo or a logo. Remove a box with its × button.
- Tap Redact and check. The tool builds the new PDF, re-opens it and shows how many redacted strings it could still find.
- When it reads 0 recoverable, download the redacted PDF.
Why most “redacted” PDFs are not redacted
The most common redaction mistake is drawing a black rectangle, or a highlight set to black, over the text. It looks right on screen and in print, but the words are still in the file underneath the shape. Select the area and paste, or search the PDF, and the “redacted” text comes straight back. The same goes for comment boxes, black-filled form fields and “mark for redaction” tools that were never applied.
This tool does not draw shapes over your page. It rebuilds every page you redact from pixels: the page is rendered, your boxes are painted into the image, and the new page contains only that image. Pages you did not touch are kept as they are, with their text still selectable.
When you open a file, it also looks for the classic failure: text that can be copied but is covered by something dark. If it finds any, it marks those places for real redaction and tells you.
What gets checked after export
| Check | How |
|---|---|
| Copy and search | pdf.js extracts the text of every page of the new file; none of your redacted strings may appear. |
| Raw file data | Every stream and string in the file is decoded and searched, including ones a viewer never shows. |
| Boxes | Each redacted area is rendered and must be solid black. |
| Text under boxes | No text may sit under a dark area anywhere in the file. |
| Hidden data | No metadata, comments, form fields, attachments, scripts or bookmarks are left. |
If a redacted string still appears somewhere you did not redact, for example a name you boxed on page 2 that is also printed on page 5, the check says where, and one tap redacts every match and checks again.
Honest limits
- Redacted pages become images, so their remaining text is not selectable. That is the price of certainty; pages without redactions are unaffected.
- Search cannot see text inside scanned images. Draw boxes over scans instead. For a photo, scan or e-Aadhaar PDF of an ID card, mask the Aadhaar, PAN or CNIC number instead: it reads the card with OCR and finds the number for you.
- Box positions for search matches are estimated from the font and padded. Look over each box before you export; you can draw an extra box over anything that is not fully covered.
- Encrypted PDFs are rebuilt with every page as an image, and the result has no password.
More guides: black out text in a PDF, remove text from a PDF permanently, redact a bank statement before sending it and check whether a PDF was redacted properly.
Questions people ask
Is my PDF uploaded?
No. The PDF is opened by pdf.js and pdf-lib running in your own browser tab, and the redacted file is built there too. Nothing is sent to a server and there is no account. The only files the page fetches are its own code and font files from this site.
How is this different from drawing a black box?
A black box drawn in most PDF editors is a shape on top of the page. The text under it is still in the file, so anyone can select it, copy it, or search for it. Here, every page with a box is rendered to pixels, the boxes are burned into those pixels, and the page is rebuilt from that image alone. The original text, fonts, vector drawings and images of that page are not carried over, so there is nothing underneath to recover.
What does “Verified: 0 of 14 redacted strings recoverable” mean?
After building the file, the tool opens it again as a fresh document and tries to get every redacted string back: it extracts the text of every page, decodes and searches every stream and string in the raw file, renders each redacted area to check it is solid black, and looks for any text sitting under a dark box. The number is how many of your redacted strings it could still find. You only download after you have seen that check.
Can I still select text after redacting?
On pages with no redactions, yes: those pages are copied across unchanged. On pages you redacted, no: they become images (at 200 dpi by default), which is what makes the removal certain. If you need searchable text on those pages, run OCR on the redacted file afterwards; the redacted parts will stay black because they are black in the pixels.
What hidden data is removed?
Every export drops document info (title, author, subject, keywords, creator, producer, dates), XMP metadata, comments and sticky notes, links, form fields, attachments, JavaScript, bookmarks, the tag structure, page thumbnails and private app data. Filled-in form values and stamps that print on the page are kept as part of the page, so the document still looks the same and those values can be redacted too.
Does it find text in scanned PDFs?
No. A scan is a picture of text, so search and patterns cannot see it. Draw boxes over the parts you want removed instead; the pixels under each box are burned black the same way.
Which patterns can it find?
Email addresses, phone numbers (US and international formats), SSN-like numbers (123-45-6789 or 123 45 6789), account and card numbers (long digit runs, cards in groups of four, IBANs, masked numbers like ****1234) and dates (04/12/1987, 2026-11-03, Nov 3, 2026, 3 November 2026). You can also search for any word, name or phrase, with match case and whole word options.
Is it really free?
Yes, with no page limit, no daily limit and no watermark. It runs on your device, so it costs us nothing per use.