Redact a PDF
Runs on your device- True redaction
- Search text to redact
- Works on phones
- Files never uploaded
or drop it here · up to 100 MB
Processed on your device. Nothing is uploaded.
Draw boxes over sensitive content or search for words to black out. The content underneath is destroyed, not just hidden behind a shape.
How it works
How to redact a PDF
- 1
Open the PDF
Drop the PDF onto the page or tap Choose file. Enter the password first if it has one.
- 2
Mark what to hide
Drag across the page to draw a black box, or type a name or number into Find text to add a box over every match.
- 3
Review every page
Use the page buttons to check each box. Tap the cross on a box or Delete in the list to remove one you don't need.
- 4
Redact and download
Press Redact. Pages with boxes are flattened into images with the areas painted black, and the new PDF downloads.
Security
About Redact PDF
Last updated
Is a black box over text enough to redact a PDF?
No. A black rectangle drawn in a PDF editor only sits on top of the text, and the words underneath can still be copied, searched or revealed by deleting the shape. Court filings and government reports have leaked names and figures exactly this way. Highlighting in black, changing the font colour to match the background and covering text with a comment all have the same flaw.
Real redaction removes the content itself. This tool does that by replacing each marked page with a flat image in which the covered areas are solid black pixels. There is no text layer on that page left to copy, and the original page objects are not kept anywhere in the new file, not even as unreferenced leftovers that a forensic tool could dig out.
What does this redaction actually remove from the file?
Everything on a marked page is removed and replaced by a single JPEG image: text, vector drawings, embedded images, annotations, links and form fields. The page is rendered with pdf.js at 200 DPI, each box is painted opaque black with one extra pixel on every side so no anti-aliased edge of a letter shows, and the picture is saved at the original page size.
The output is a brand-new document, not an edited copy. In the PDF format described by ISO 32000-2, a saved file can hold objects that no page refers to any more, so editing in place risks leaving the old content behind. Here only the needed pages are copied into an empty file. Marked pages are also emptied in the source first, so an untouched page that links to a redacted one cannot drag its original content along.
| Part of the document | After redaction |
|---|---|
| Text, vectors and images under a box | Removed; replaced by black pixels |
| Other content on a page with a box | Kept as a picture at 200 DPI; no longer selectable |
| Links, comments and form fields on a page with a box | Removed |
| Pages without a box | Copied unchanged, text still selectable |
| Bookmarks, attachments and the form field tree | Not carried over |
| Info dictionary, XMP metadata and app-private data | Removed by default |
| Metadata when the option is unticked | Only title, author, subject and keywords are kept |
What should be redacted before sharing a document?
Redact any detail the recipient does not need for the purpose at hand, which usually means identifiers, contact details and figures. Before sending a document to a landlord, employer, court, insurer or journalist, read it once with that single question in mind and mark everything that fails the test. The most common candidates are:
- Account, card and policy numbers on statements
- National ID, passport and tax numbers
- Home addresses, phone numbers and signatures
- Names of children, patients or witnesses
- Salary figures and pricing in shared contracts
How do I make sure nothing is missed?
Use Find text for anything that repeats, then review every page by eye, because a search can miss words split across lines or written in unusual fonts. Try variations such as an account number with and without spaces, or a first name on its own. Draw boxes slightly larger than the text so the tops of tall letters and the tails of others don't show.
Check headers, footers, stamps and signatures, and remember that the same fact can appear in a chart label, a table or a photo. After downloading, open the result and try selecting or searching for the redacted words to confirm nothing remains. Keep the unredacted original somewhere private, because redaction can't be undone in the new file.
What are the limits of image-based redaction?
The main limit is that redacted pages become images, so they are larger than text pages, can't be edited as text and are not searchable until OCR is run. Long documents with boxes on hundreds of pages take a while on phones because each page is rendered one at a time. For a smaller file afterwards use Compress PDF, and to add a password before sending use Protect PDF.
Pages left untouched keep their text, links and quality exactly as they were. If a document needs only a few words hidden on a long text page, that whole page still becomes an image, so consider whether sending only the relevant pages with Split PDF suits the recipient better. Files up to 100 MB are supported, and a password-protected PDF is saved without its password once redacted.
FAQ
Redact PDF questions
Is the redacted text really gone, or just covered?
It is gone. Each page with a box is rendered to an image, the boxes are painted onto that image, and the original page is replaced. No hidden text, vector shapes or images from under the boxes remain in the file.
Why can't I select text on the redacted pages any more?
Because those pages become images at 200 DPI, which is what makes the redaction permanent. Pages without boxes stay unchanged, so their text is still selectable. Run OCR PDF afterwards if you need searchable text on the redacted pages again.
Can I redact a PDF on my phone?
Yes. Switch to Draw box and drag with a finger to mark an area, or use the text search, which is often quicker on a small screen. Switch back to Scroll to move around the page without drawing.
Does the search find text in scanned documents?
Only if the scan has a text layer. Plain scans are pictures of text, so there is nothing to search. Run OCR PDF first to add a text layer, or draw the boxes by hand over the scanned words.
Where is the document processed while I redact it?
Entirely in the browser. Rendering, text search and building the redacted PDF all run on your device and the file is never sent to a server, which matters for exactly the kind of documents people redact.
Is hidden information removed too?
Yes, by default. Document metadata such as title, author and software names is removed unless the option is unticked. The redacted file is rebuilt from scratch, so bookmarks, attachments and the form field tree are not carried over either.
Can a redaction be undone?
No. The downloaded file holds black pixels where the content used to be, and nothing of the original page is stored in it. Keep the unredacted original somewhere private if you may need the full text again.
Keep going
Related tools
Got a PDF to fix?
Free, no account, no watermark. Most tools run on your device, so files never leave it.
Browse all tools