Redaction means removing content, not covering it with a comment
A black rectangle added by an ordinary annotation tool can leave the words underneath intact. That is why this tool does not copy the original PDF pages into its output. It renders each visible page, overwrites your selected regions in the image, and puts only those finished images into a newly created PDF.
The result is image-only. Someone cannot select the original words beneath those boxes from the rebuilt PDF because that text layer is not included. However, anything you did not cover remains visible, and surrounding context may still reveal information. Review the exported document, not only the editing screen.
How to redact a PDF in UVs Tool Hub
- Open a PDF from your device. Password-protected files require the password and permission to edit.
- Draw boxes over private text, numbers, photos or signatures. Switch to Select to move and resize a box.
- Use Find text for literal words, phrases or possible email, phone and card-number patterns. Review candidates before marking them.
- Check each page. Repeated areas can be copied to a page range, but their placement must still be reviewed.
- Choose the output resolution, apply the redactions, preview the actual generated PDF and download that new file.
Useful for documents you need to share selectively
A support request may need an invoice without a home address. A sample contract may need a person’s name and signature removed. A report may contain a barcode, photograph or reference number that should not reach every reader. Manual area selection works with these visual elements, even when they cannot be searched as text.
Do not use a narrow text box when an entire address block, QR code or picture needs removal. Mark all of it. For repeated headers or account details, copy a selected area to other pages and inspect each copy. Different page sizes can change what occupies that relative position.
Search results are candidates—not an automatic privacy guarantee
The search works with the PDF’s selectable text layer. It can find typed words and phrases, and it offers built-in patterns for likely email addresses, phone-like numbers and card-like numbers with a checksum. These patterns can produce false positives and miss unusual formats. Text split into unusual encoded fragments or stored inside an image may not be found.
Search redaction covers complete matching text runs rather than guessing individual glyph widths. That can cover more text than the phrase you entered. Inspect and resize the proposed boxes where needed. Scanned PDFs do not get OCR in this tool; draw regions manually on the page.
What the rebuilt file keeps—and what it drops
The exported PDF keeps the visible unredacted page appearance at your chosen resolution and preserves each displayed page’s physical dimensions. Choose 150 DPI for a smaller file, 200 DPI for a balance or 300 DPI for more detailed printing. Larger pages and more pages consume more device memory.
It does not preserve searchable text, accessibility tags, interactive forms, live links, original comments, attached files, hidden layers or the original metadata. Existing digital signatures are not valid on the new document, and original password protection is not carried forward. Some visible annotation appearances may be rasterised into the page. Review them like any other visible content.
Free without sending your PDF to a converter
Your file is opened and processed in this browser tab using locally hosted libraries. The redactor does not send the PDF to our server or an outside conversion provider. It does not store a recovery copy of this document. Closing the page clears the editor session; the original file and any downloads on your device remain your responsibility.
The website still uses the analytics and feedback features described in our Privacy Policy. Do not attach a sensitive PDF or a screenshot showing private details to a feedback message.
Before sharing a sensitive document
Review every page of the rebuilt PDF. Check small lettering around a box, repeated occurrences, unmarked pictures, signatures and machine-readable codes. Do not assume a clean search result means the PDF contains nothing private. The process report describes the rebuild and its structure checks; it does not certify legal compliance or that every sensitive fact was selected.
For documents with serious confidentiality requirements, follow your organisation’s approved review process and get a second person to check the release copy. Use the source application to remove information before creating the PDF when that is the more appropriate route.
Related document tools
Use Split PDF to work on fewer pages, PDF to JPG / PNG for page images, or Merge PDF to combine finished release copies. Keep unredacted originals separate so they are not accidentally merged back into the file you plan to share.
Review the output, not only the marks
After export, the tool reopens the generated PDF and checks its structure and the opaque interiors of marked areas. You then review the actual exported pages before download. This can detect technical coverage problems; it cannot decide whether you marked every private detail. Keep the original separate from the redacted copy.
Save a marking plan without saving the PDF here
A downloadable marking plan records box positions and export settings. A SHA-256 fingerprint binds it to the exact PDF bytes, so it cannot silently be applied to a different document. Plans are not automatic backups and contain no original PDF pages. Keep the original document and plan private.
Frequently asked questions
Is this PDF redactor free without signup?
Yes. Manual marking, search candidates, repeated areas, output review and downloads are free without an account. File, page and processing limits protect your device rather than acting as paid-credit gates.
Can the text under a black box be selected from the output?
The output contains newly encoded page images, not the original text objects. Selected regions are overwritten before encoding. Unmarked content remains visible, and context may still reveal information. Only share the rebuilt file, not the original.
Does this work with scanned PDFs?
Yes, you can manually draw areas over a scanned page. The text finder only searches selectable PDF text and does not run OCR on scans.
Can I repeat a redaction across pages?
Yes. Select a box and use Repeat area with All or a page range such as 1-3,5. The relative position is copied, so review each destination page.
Will my PDF remain searchable?
No. Every exported page is an image. This also removes accessibility tags, interactive fields and live links. Use an approved object-level redaction workflow when preserving searchable accessible text is essential.
Are my files uploaded or saved automatically?
No document is uploaded by the redactor and no recovery copy is stored. Editing happens in memory in the current browser. Keep downloads private, and avoid sending confidential content in feedback screenshots.
Can I undo a redaction?
You can undo or redo marks before creating the output. In the rebuilt PDF the selected source pixels are not retained. To change your selection later, return to the original file and create another redacted copy.
Does the check guarantee that my document is safe to share?
No. The structural check confirms the rebuilt page count and dimensions and checks for selectable text, annotations, attachments and document actions. It cannot know whether you marked every private detail or whether context still reveals it.
Can I save my redaction marks for another session?
Yes. Save a marking-plan JSON file, keep the exact original PDF, and reopen both later. A fingerprint prevents applying the plan to a different PDF. Plans contain box positions and settings, not the original document or search results. They are not encrypted.
Why is the final download button locked?
After the tool verifies its rebuilt output, open Review exported PDF and confirm each rendered page. Then tick the final review checkbox. New mark or output-setting changes invalidate the result and require another export and review.