PDF Redaction Studio
Render a PDF to page images in your browser, then draw or enter the regions to cover. The output is a new PDF containing only one masked image per page; original PDF page objects are never copied. It checks the output text layer, image object count, annotations, forms and attachment/script name tree. This is a bounded full-page raster workflow, not surgical deletion of only the selected native PDF objects.
Key features
- Mark page regions by mouse/touch drag or keyboard-friendly percentage fields
- Rasterize each source page to PNG, paint opaque mask pixels, and create a fresh image-only PDF
- Check zero extracted text, one image object per page, and no annotations, forms or attachment name tree
- Download the redacted PDF and a JSON inspection report
- Reject encrypted and oversized PDFs; use a local worker without remote loading
How to use
- Choose a password-free PDF no larger than 8 MB and 5 pages, or open the built-in example.
- Drag a generous box over every sensitive region; keyboard users can enter page, X/Y position, width and height as percentages.
- Review the black preview and mask list; remove and add masks until they cover the complete content.
- Generate a new image-only PDF and inspect the automated object and text-layer checks.
- Save the PDF and report, then zoom into every output page in a PDF viewer to inspect boundaries and omissions.
Use cases
- Produce a new raster PDF where visible personal details are covered
- Check that selectable source text and attachments or annotations did not carry into the output
- Conservative redaction of a small PDF when loss of searchability and vectors is acceptable
Frequently asked questions
Is this just a black box drawn on top of the original PDF?
No. It renders each source page to an image, paints the mask pixels, and creates a fresh PDF without copying original page objects. It then checks that no text layer, attachments or annotations remain.
Will text quality and accessibility remain intact?
No. Every page becomes a PNG image. Searchable/copyable text, vector sharpness, screen-reader structure, form fields and links are lost. Resolution is bounded, and file size may increase.
Does a passed inspection prove every secret is gone?
It checks the new PDF structure for a text layer and carried source objects. It cannot catch visible text outside a mask or a mask drawn too narrowly. Inspect every output page yourself.
Can it handle password-protected or very large PDFs?
Encrypted PDFs are unsupported. Limits are 8 MB input, 5 pages, 40 MB output, 2400 pixels on one page side and 4 million pixels per page. Use a trusted dedicated editor for larger files.
Privacy
Files are read and processed within this browser tab. PDF bytes are not uploaded to our server or an external service. The PDF.js worker, fonts, CMaps and WASM assets are served from this site. After saving, inspect the output visually to confirm every intended region is covered.
Comments & questions