Blur or pixelate sensitive areas
Drag boxes over faces, license plates, addresses or screen contents and the pixels underneath are genuinely destroyed, not just hidden. The redacted image is safe to post. Nothing is uploaded.
Drop an image to redact
JPG, PNG, WEBP, then drag to draw boxes over sensitive areas
The detail is gone, not hidden
A face pixelated with this tool. Drag the handle: the redacted side cannot be turned back into a face.
Original
Pixelated
Because the pixels in that box are overwritten with a coarse mosaic, there is no original face left in the saved file to recover.
Pixelate, blur, or a solid box
Pixelate, for real privacy
A coarse mosaic genuinely averages the detail away. This is the safest option for faces, plates and anything that must not be identifiable.
Blur, for a softer look
Good when you want to de-emphasise something rather than fully hide it. Be careful: a light blur can sometimes be partly reversed, so go strong if it is sensitive.
Solid box, for zero doubt
A black rectangle leaves nothing at all. Use it for text like account numbers where even a hint of the shape would give it away.
Not the same as a blur in a phone app
Many editors and CSS blurs only cover the pixels on screen while keeping the original data in the file, so the hidden part can resurface when the image is re-processed. This tool rewrites the actual pixels in the box and exports a fresh image, so what you cannot see is genuinely not there. It all happens in your browser, so the original photo never leaves your device either.
Frequently asked questions
Blur smooths pixels together so features become fuzzy. Pixelate replaces blocks with their average color, a hard mosaic. For privacy, pixelate at a large block size is the safer choice, because a strong blur can sometimes be partly reconstructed by AI tools while a coarse mosaic genuinely throws the detail away.
Yes. The tool rewrites the actual pixels in that region, so the original detail is not recoverable from the file. That is different from a CSS or app blur that only hides the pixels visually while leaving them in the data.
Yes. Drag to draw a box, then drag again for the next one, up to 20 regions per image.
Enough that the face is an unrecognizable mosaic, not just soft. For a face filling roughly 200 pixels, a block size around 16 to 24 is a safe starting point. When in doubt, go stronger.
No. Everything runs in your browser. The original photo never leaves your device, and the redacted version is generated locally too.