Skip to content

PDF · Redaction

Redact a PDF: the masked text is removed from the file

A black box drawn on a PDF only hides the text from view. The text stays in the file, and a simple copy and paste can bring it back. ONYRI Sanitize is a PDF redaction tool that works differently. It finds the sensitive data, you mask the rest by hand, and it exports a flattened copy with no text under the bars. Everything runs in your browser.
Redact a PDFSee plans and pricing

Runs 100% in your browser. Your file is never uploaded.

What text and data you can redact in a PDF

  • A word or a sentence: select it, then mask every occurrence or just one
  • A term that comes back often: search for it, then mask all the results in one click
  • A signature, a logo or a photo: draw an area over it
  • Names of people, found in body text, headers and signatures
  • Email addresses, phone numbers and postal addresses
  • ID, passport and Social Security numbers, and birth dates
  • IBANs, card numbers and company identifiers
  • Values typed into the fields of a filled-in PDF form

How to redact a PDF in three steps

  1. Drop the PDF

    Create your account, then drag the file onto the page or pick it from your disk. ONYRI reads the text of each page in your browser and finds the sensitive data. A scanned page is read by on-device OCR. Nothing from the document is uploaded.

  2. Choose what disappears

    Detected values are already masked on the page. Untick anything that should stay readable. For the rest, use “Select text” or “Draw an area”, or search for a term and mask all the results. Click a mask and choose “Don’t mask” to take it back.

  3. Download the flattened copy

    You get a new PDF with the same pages. Each page is redrawn as an image, black bars included. No text sits underneath and no rectangle can be moved. To check, search the new file for a word you masked: it cannot be found.

Why a black box does not redact a PDF

A PDF keeps its text as real characters. A rectangle drawn in an editor is a separate object laid on top. The text is still underneath. Anyone can select it, copy it or search for it. Sometimes the rectangle can even be moved.

ONYRI flattens the file. Each page becomes an image with the black bars inside it, so no text layer is left to copy. The new PDF is also built from scratch. It does not copy the title, the author, the bookmarks or the comment objects of the original. What is visible on a page, such as a text-box comment, is still drawn in the image, so check each page.

The document stays on your device. The PDF is read, masked and exported inside your browser tab. You can check it in the Network tab of the developer tools: no request carries your file. Only counters and codes, with no content, are recorded: pages, masks by type, the country and language recognised, a token of the file name.

Honest limits

What you should know before you send the file

  • Automatic detection is not exhaustive. Read every page of the preview before you download, and mask by hand what is missing. A lone name, an initial or a nickname can slip through. Text inside comments is not read.
  • Scanned pages are read by on-device OCR, page by page. It is slower, and quality depends on the scan. Handwriting, signatures and photos are not read as text, so draw an area over them. If a page stays unreadable, the download stays locked until you tick a box confirming that you reviewed the page and masked what was needed yourself.
  • The exported PDF is made of page images. It holds no text of its own, and nothing can be restored. Keep your original. The downloaded file reuses the original name followed by “-anonymized”, so rename it if that name contains sensitive data.
  • Redaction reduces exposure. On its own, it does not make you GDPR compliant. A blacked-out document can still point to someone through context, such as a job title, a date or a place.

Redacting a PDF: frequent questions

How do I redact a PDF so the text is not left underneath?

A black box is only a shape laid on the page, so the text stays in the file. To remove it for real, you must delete it or turn the page into an image with the bars inside. ONYRI does the second.

How do I check that a PDF is really redacted?

Open the final file in a different PDF reader. Search for a word you masked. Then select all, copy and paste into a plain text editor. The masked words must not appear. ONYRI's export has no text layer. Some readers, such as Chrome, read the page image with OCR, so visible words may still paste. Masked words never can.

Can I redact a PDF on a Mac without installing anything?

Yes. Open ONYRI Sanitize in your browser, on a Mac as on a Windows PC. There is nothing to install and the file is not uploaded. ONYRI's servers handle your account, your plan and usage counters. They never receive the document or its content.

Can I redact just one sentence, a signature or an area, and take it back?

Yes. “Select text” masks a word or a sentence, everywhere or in one place only. “Draw an area” covers a signature, a logo or a photo. Before you download, click a mask to remove it, or use Undo to cancel your last action. In the final PDF, nothing can be removed.

Does it work on a scanned PDF or a filled-in form?

Yes. A page with no text layer is read by on-device OCR. Values typed into form fields are read too. If a page stays unreadable, you mask it yourself and confirm your review before downloading.

How many documents and pages can I redact?

You need an account. The Free plan allows 1 document a day and 5 pages per document, in black-marker mode (see pricing). The 7-day pass or Pro removes those limits and adds token mode.

Your document, made confidential in seconds.

Drop a file, check the preview, download the masked copy. The file never leaves your device.