Overview
What this tool does
Real PDF redaction permanently removes content from a document, unlike most "redaction" tools that just draw a black box on top (the underlying text is still recoverable from the PDF byte stream by anyone who opens it in a PDF editor). This tool re-renders any page that has redaction marks into an image with the black bars baked into the pixels, then assembles a new PDF. Pages WITHOUT any redactions are copied verbatim, so their text stays selectable. The trade-off is that pages WITH redactions become image-only (text on those pages is no longer selectable / searchable). For legal, HR, or compliance use where leakage is a serious risk, this is the right trade-off. 100% client-side via pdfjs-dist and pdf-lib; no upload.
How to
Use it in 3 steps
- Drop your PDF on the page (or click 'Pick PDF').
- On the page preview, drag rectangles over any text, signatures, or regions you want redacted. The bars appear as you drag.
- Use 'Prev' / 'Next' to navigate pages and add more redactions. The total mark count is shown above the page.
- Hover any black bar and click the red X to remove that one redaction. 'Clear page' wipes all redactions on the current page; 'Clear all' wipes all.
- Click 'Apply & download' when ready. The tool rasterises every page that has marks, paints the black rectangles into the pixels, and assembles a fresh PDF for download.
FAQ
Frequently asked questions
▶Is the redacted content really gone?
Yes. The text under each black bar is replaced with solid-black pixels in a rendered image; the original text data is not embedded in the output PDF at all on those pages. There is no way to recover it from the file (short of OCR-ing the black box, which obviously fails).
▶Why does the rest of the page lose selectable text too?
Because each affected page is replaced as a single image. We could in theory rasterise just the redacted area and keep the rest as vector text, but reliably surgically removing only the affected text runs from the underlying PDF content stream is complex and error-prone. Rasterising the whole page guarantees the redacted content is gone, with the cost being that you can't copy-paste from those specific pages anymore. Unredacted pages keep their vector text intact.
▶Is my PDF uploaded?
No. Rendering uses pdfjs-dist in your browser, assembly uses pdf-lib in your browser. The file never reaches a server.
▶What about images that contained text inside the redacted area?
Pixels in the redacted rectangle are overwritten with solid black, including any image content underneath. Even if the redacted box covers part of a photo or scanned document, those pixels are GONE in the output.
▶Why does the output file size grow?
Image-based pages compress as JPEG, which can be larger than the vector original (vector text is extremely compact). The DPI we render at (144) balances visual quality against file size. For very large documents, expect the redacted PDF to be larger than the source.
▶What's the difference between this and the PDF Editor's redact tool?
The PDF Editor's redact tool draws a black box on top as an annotation, which looks redacted but the original content is still embedded in the file and recoverable. This tool is for cases where the leakage risk is real and the content must actually be removed.
▶Can I redact across multiple PDFs?
Not in one pass. Process each PDF separately. Each output is a fresh standalone file.