Redact PDF
Mark text by search, by pattern (emails, phone numbers, card numbers and more) or by drawing boxes. What’s under each box is deleted from the file, not covered up. Then the new file is checked: its text is re-read, and the area under each box is drawn again without the box and checked to be blank. A page where something still shows is redrawn as an image with the boxes burned in.
What Redact PDF does
Redact PDF permanently removes sensitive content from a PDF: names, account numbers, addresses or any area you mark. Drawing a black box over text in an ordinary editor leaves the text in the file, where it can be copied out. This tool deletes the text, image pixels and drawings under each box and then draws the box.
After redacting, the tool checks the new file twice. It re-reads the text to make sure none is left inside a box, and it draws each box's area again without the box to make sure no image or drawing still shows there. If something does, that page is turned into an image with the box burned in, and checked again.
How to redact a PDF
- Drop a PDF on the box above.
- Mark what to remove: search under Find text, pick patterns under Personal info, or turn on Draw boxes and drag over any area.
- Check the list of marks and untick any you want to keep, then select Redact. Read the verification result before you share the file.
What it finds and removes
| Find text | Every match of a word or phrase. Options: Match case, and Regular expression (JavaScript syntax). |
|---|---|
| Personal info | Email addresses, phone numbers, card numbers, IBANs, US Social Security numbers and dates. |
| Draw boxes | Drag over any area of a page, including images and drawings. |
| Box color | Black or white. |
| Removal | Text, image pixels and vector drawings inside each box are deleted, then the box is drawn in their place. |
| Remove hidden data | On by default. Also removes document properties, XMP metadata, attachments, JavaScript, and form fields or comments under the boxes. |
| Verification | The new file's text is read again, and each box's area is drawn without the box to check it's blank. If anything is still there after that, you're told not to share the file and shown where. If this browser can't draw the check, the result says images and drawings weren't checked. |
Limits
- Automatic detection finds patterns, not meaning. Read the document and mark anything it missed.
- The checks cover the areas inside your boxes. Anything you didn't mark, such as the same name on another page, stays in the file.
- If content under a box can't be removed any other way, or still shows after removal, that page is turned into an image. Its text can't be selected or searched until you run OCR PDF on it.
- Redaction can't be undone in the output file. Keep your original somewhere safe.
Where your file goes
Redaction runs in this browser tab. The PDF and the text you search for aren't uploaded, which matters most for exactly the documents you redact.
If you're signed in, the result panel also offers Save to My documents. Only that button uploads the file, to your account's storage.
Questions
Is a black box over text enough to redact a PDF?
No. A box drawn with an ordinary editor only covers the text; it can still be selected, copied or extracted. This tool deletes the content under the box from the file.
How do I know the redaction worked?
After redacting, the tool re-reads the new file's text under each box, and draws each box's area without the box to check that no image or drawing still shows. A Verified message lists both checks; otherwise you're told not to share the file and shown what's left. Still look over the pages for anything you didn't mark.
Can it find personal data automatically?
It finds patterns: email addresses, phone numbers, card numbers, IBANs, US Social Security numbers and dates. Patterns can miss things or catch too much, so check each mark. Names and addresses need Find text or a box.
Can I undo a redaction?
Not in the redacted file; the content is gone. Your original file isn't changed, so keep it if you may need it.
Does it work on scanned PDFs?
Boxes you draw remove the image pixels under them. Search and personal-data detection need text, so run OCR PDF first on a scan.