Redact PDF

Open a PDF and every page appears in a scrollable viewer. Mark what has to go: type the words to look for, tick the kinds of personal information to find, or drag boxes over anything on the page. Review the marks, then run the redaction. You get a new copy of the document; the original on your device is not changed.

The difference from a simple cover is the whole point of the tool. Under each box, the text is deleted, the covered pixels of pictures are blanked and touching lines and shapes are removed. Then the file is cleaned and searched again, so you can see with your own eyes that the marked words cannot be found anymore.

There are two very different ways to hide something in a PDF. The old way is to draw a black rectangle on top of the words. On screen and on paper the result looks safe, but the words are still stored in the file underneath. Anyone can select the area and copy the text, search for it, or remove the rectangle in an editor. This mistake has exposed confidential details in published documents more than once.

Real redaction removes the content itself. The tool sends the PDF to our server, where PyMuPDF, a well-known PDF library, places a redaction mark on each box and then applies it. Applying deletes every letter whose position falls under a box, cuts out the covered pixels from pictures, and removes the lines and shapes the box touches. Only then is the box painted in the colour you picked, so the black area is a sign that something was removed, not the only thing hiding it.

After that the file is cleaned: unused objects are dropped, page content is rewritten, page thumbnails are removed and, if you leave the default options on, the document details, hidden text, scripts and attached files go too. The result is a normal PDF that opens in any reader, with the marked content missing rather than hidden.

To redact pdf text such as a person's name in a 40-page report, type the name, press Find and every place it appears is marked at once. That is far quicker and safer than hunting for each mention by eye, and it does not miss the one copy hidden in a footnote or a table.

The patterns are written to be practical rather than perfect. Card numbers are checked with the Luhn rule, the checksum every payment card number follows, so random long numbers are not marked as cards. Phone numbers need at least seven digits. Long numbers catches IDs, account numbers and case numbers with spaces or dashes inside. Dates covers forms such as 12/05/2026, 2026-05-12 and 5 May 2026.

The search works line by line, the way the text is stored in the file. A phrase that is broken across two lines will not be found as one match; search for each part, or add a box by hand. Up to 5,000 matches are listed, which is plenty for nearly any document.

The Matches to review panel groups the matches by kind: Your words, Your pattern, Email addresses and so on, each with a count such as 12/14 that shows how many are still marked. Untick a group to skip it all, or untick single matches. Click the text of a match to jump to its page.

On the page, a marked match has a red box and a skipped one has a thin dashed outline. Click a box to switch between the two. That is how you keep the company phone number on the letterhead while removing a private one in the body of the letter, or keep the date of a report but remove a date of birth.

Some things cannot be found by searching: a signature, a photo, a stamp, a handwritten note or a part of a scanned page. For those, use the Mark by hand card. With Draw a box selected, drag on the page to cover any area. With Mark text selected, drag across words or click a single word and the box snaps exactly to the text, so you do not cut letters in half or catch the line below.

To redact whole pages, type them in the Redact whole pages field, such as 3, 7-9, or tick Redact whole page above any page. The whole page is then cleared and filled with your box colour, while the page itself stays in the document so the page numbers still match. If you would rather drop the pages altogether, use Delete Pages instead.

The bar above the pages shows how many areas are marked. Undo removes the last box, Clear marks removes them all, and the small cross on a box removes just that one. Tick Preview as redacted to see the pages with solid boxes in your colour, the way they will look in the file. Zoom from 60% to 200% for exact work on small print.

Under Look of the boxes, choose Black, White or Grey, or pick any colour with the colour box. Black is the classic look and tells every reader that something was removed. To redact pdf with white box fills is just as safe: the content is still removed, but the page looks clean, which suits forms, printed handouts and documents that will be filled in again.

Text on the box can be None, REDACTED or Your text, up to 60 characters, for example Confidential or an exemption code such as (b)(6). The words are centred in the box, in white on dark colours and in black on light ones, and they are only added when a box is tall enough to read them.

Scans, photos, logos and screenshots are pictures inside the PDF. Pictures under a box has three choices. Blank the covered part, the default, cuts out only the pixels under the box and leaves the rest of the picture as it was. Remove the whole picture deletes every picture the box touches, which is the strict choice when even the edge of a photo should not stay. Leave pictures keeps pictures untouched and only works on text and drawings.

Drawings under a box has one switch: Remove lines and shapes touched by a box, on by default. Many signatures made in a PDF program, charts and drawn marks are made of lines and shapes rather than pictures, so this setting makes sure they do not stay behind under the box. Turn it off if a box touches a table border or a frame you want to keep.

A PDF can carry far more than what you see on the page. Under Hidden data in the extra options you choose what else is stripped out. Remove author, title and other document details clears the information fields and the XML metadata, which often hold a person's name, the program used and the original file title. Remove hidden text deletes text that is in the file but not shown, a common trap in scanned PDFs and old drafts.

Remove JavaScript takes out scripts, and Remove attached files deletes files embedded in the PDF. These four are on by default. Remove comments and markup, off by default, deletes notes, highlights and other review marks but keeps form fields and links. Remove links, also off, removes clickable links, which can contain an e-mail address or an internal web address.

For the file name, Add -redacted to the name is on by default, so the copy is easy to tell from the original. For a quick clean-up of document details without any redaction, Remove Metadata is made for that job.

When the file is saved, the server opens it again and runs the same search you used with Find: the same words, the same case and whole-word settings, the same patterns and your own pattern. The result card then tells you plainly. If nothing is left, you see the line that none of the marked words or patterns can be found in the new file.

If something is still there, the card says how many matches are left, for example because you skipped them on purpose or because a copy sits on a line the first search did not catch. Look again, add boxes and redact once more. You also see how many areas were removed on how many pages, the new file size and small previews of up to four pages that were changed.

The check covers what you searched for. Boxes you drew by hand over a signature or a photo have no words to search, so look at the previews and, if you like, open the downloaded copy and try to select the area. Nothing should be selectable there.

A scanned page is a photo of paper with no text layer, so Find cannot read it. After a search, the tool tells you how many pages have no text layer. On those pages, draw boxes by hand over the parts to remove: with Blank the covered part, the pixels under each box are cut out of the scan for good.

If you would like to search a scan by its words, make the text readable first with PDF to Text, which uses OCR to read the words. OCR can misread names and numbers, so check the result before you trust a search, and still look over every page before you share the file.

A PDF with an open password cannot be read or redacted without it. When you open one, a password box appears: type the password and press Unlock. The file and the password are sent once over an encrypted connection to unlock it, and the unlocked copy comes back for marking. The redacted copy you download has no password.

If you only need the password removed, Unlock PDF does that on its own. After redacting, Protect PDF adds a new password before you send the file on.

Redaction is a daily task wherever personal data is shared. Under privacy laws such as the GDPR in Europe, organisations often remove the personal data of other people before they hand out a document. Health records shared under rules like HIPAA in the United States are cleared of patient details, and public bodies answering FOIA and other freedom-of-information requests black out names and protected passages, often with an exemption code on the box.

This tool gives you the means to do that properly: content removed, not covered, with a check afterwards. What has to be removed in your case, and why, depends on your country, your sector and your document, so follow your organisation's rules or ask a legal adviser. This page is general information, not legal advice.

Every option on this page is free: Find and redact with all ten patterns and your own pattern, boxes by hand, whole pages, box colours and labels, picture and drawing options, hidden data clean-up and the search check. You can redact pdf online free with no daily limit on files and nothing held back for an upgrade.

The only limit is size: PDFs up to 100 MB can be opened. For a bigger file, Compress PDF makes it smaller or Split PDF cuts it into parts you can redact one by one.

There is no account, no e-mail address and no trial. Open the page, add the file, mark and download. The redacted copy carries no watermark and no line from us, so you can send it on as your own document.

Anyone who wants to redact pdf free no sign up can do it in under a minute: open the page, select the file, press Find or draw a box, then press the button under the settings and download.

Many free tools that promise redaction only lay black shapes over the page. This redact pdf tool free of charge does the real thing: the text under each box is deleted, the pixels are cut out and the drawings are removed, then the file is checked by searching it again.

The work is done by PyMuPDF on our server, the same kind of library professional PDF software is built on. No part of the job is left to a picture laid on top, and the report tells you how many areas were removed.

Here is exactly what happens. The pages are shown in your browser, and your boxes are drawn there. When you press Find, and again when you run the redaction, the PDF is uploaded to our server over an encrypted HTTPS connection, because the removal is done there by PyMuPDF. The file is deleted from the server right after the result is sent back to you. We do not keep a copy, look at it or use it for anything else.

The original on your computer or phone is never changed: you always download a new copy. When you redact pdf file copies of confidential papers, keep the original in a safe place and share only the redacted copy. Closing the tab clears the file from the page.

Full desktop PDF suites, Acrobat Pro among them, have good redaction tools, but they need a paid licence and an install. If you only redact a document now and then, you can redact pdf without adobe software or any other paid program on this page. Any modern browser is enough: Chrome, Edge, Firefox, Safari or Samsung Internet.

That helps on a work computer where you may not install software, on a borrowed laptop, or when a colleague needs a one-off redaction. For large legal workflows with hundreds of files a day, a desktop suite may still suit you better, as the limits below explain.

On a Mac, open the page in Safari, Chrome or Firefox and drag the PDF from Finder onto the box. Use the trackpad to drag boxes on the pages. Cmd+Z undoes the last box and Cmd+Enter starts the redaction. The copy lands in your Downloads folder.

Nothing has to be installed, and it works the same on an Intel Mac or Apple silicon. If you need to redact pdf on mac with a search for names and patterns, a list of matches to review and a search check afterwards, this page adds those steps with no app to install.

On Windows, open the page in Edge, Chrome or Firefox. Drag the file from File Explorer, click Select PDF file, or copy the file and paste it on the page with Ctrl+V. Ctrl+Z undoes a box and Ctrl+Enter runs the redaction. The copy is saved to your Downloads folder.

Windows has no built-in PDF redaction, so most people install a program for it. To redact pdf on windows with nothing to install, this page works in any browser, on any PC that can open a web page, and on Linux and Chromebooks the same way.

On an iPhone or iPad, open the page in Safari or Chrome, tap Select PDF file and choose the document from Files, iCloud Drive or another storage app. Type words under Find and redact and tap Find, or drag a finger on a page to draw a box; the page does not scroll while you draw. Tap a match to skip it.

To redact pdf on iphone comfortably, turn the phone sideways for a wider page and pick a higher zoom for small print. The Undo button and Clear marks replace the keyboard shortcuts. The redacted copy goes to Downloads in the Files app, where you can share it. On Android the steps are the same, and the copy is saved to Downloads.

Before you share the copy, take two minutes for a last look. Open it in the PDF reader you normally use, scroll through every page and make sure each box sits where it should and nothing is half covered. Then use the search box of the reader to look for two or three of the words you removed. They should not be found.

Try to select text inside one of the boxes and copy it. Nothing should come out. Look at the document properties in the reader to see that the author and title fields are empty, if you kept that option on. If something looks wrong, go back to the tool, press Change marks, fix the marks and run the redaction again; the file is still open there, so you do not have to start over.

For a set of related documents, such as the letters in one case, keep a short list of the names, numbers and terms to remove. Paste the same list into Find and redact for each file, tick the same patterns, and you get the same result every time. Your box colour, label and extra options are remembered on this device, so you only set them once.

If the documents belong together, you can also join them first with Merge PDF, redact the combined file in one pass and then cut it back into parts with Split PDF. That way one review of the matches covers every page at once.

Redaction cannot be undone in the copy: the removed text, pixels and drawings are gone from that file. Your original stays as it was, so keep it if you may need it again.

Find reads the text layer only. Scans need boxes drawn by hand or OCR first, and text inside a picture, such as a screenshot, is not found by search. Phrases broken across two lines are not found as one match.

Lines and shapes are removed when a box touches them, which can take a table border with it; turn off the drawing option if that matters. Text under a label must fit in the box, so very small boxes get no label.

The tool works on one file at a time, and every Find and redaction sends the file to our server, so a slow connection makes large files slower. It does not keep a log of what was redacted, and it does not add a digital signature.

With your redacted copy saved, Protect PDF adds a password before you send it, Compress PDF makes it small enough for e-mail and Merge PDF joins it with other documents. Split PDF takes out just the pages you want to share.

To show someone exactly what changed between the original and the redacted copy, Compare PDF puts the two side by side with the differences marked.