Redact PDF

A black box laid over text (annotation, drawing, highlight) still contains the original text underneath, a recipient can copy, search, or reveal it by removing the annotation. Plainjar's redaction rasterises the page itself: the text layer is replaced with an image where redacted areas are baked in as black pixels. No text to recover. Test it yourself: search the output for the redacted words, nothing.

On pages without redactions, yes, those stay intact. Pages with redactions are rasterised at 150 DPI: the whole page is replaced with an image with the black bars baked into it. On those image pages, text can't be selected, and search tools (Ctrl-F in Acrobat, screen readers, search indexers) won't find any text. That's by design, it's exactly how the bytes on those pages vanish. If you need the redacted PDF to be searchable again, run OCR over the result (e.g. via Acrobat or a dedicated OCR tool). Also note that image pages are slightly larger in file size than the original text pages.

No. Rendering uses pdfjs-dist in your browser, redaction uses Canvas, output PDF assembly uses pdf-lib. No server call, no log. Open the Network tab during use, apart from loading the pdfjs worker and the tool assets, nothing happens.

Everything that can identify the PDF: /Info (author/title/producer/dates), XMP, thumbnails, per-page metadata, all annotations (sticky notes, links, comments), form fields including filled values, embedded files, JavaScript actions, bookmarks, and the entire name tree. This is THE strip-everything tool.

Redaction can't load encrypted PDFs. Remove the password first with 'Remove PDF password' (coming soon, you can upvote it now to push it up the priority list). Then redaction will work.