Image Redactor
Brush over faces, number plates and private text to blur, pixelate or black them out — in your browser, without uploading anything.
Table of contents
How to blur or pixelate part of an image
- 1
Open a picture
Drop in a photo, choose a file, or paste a screenshot straight from the clipboard. It is read into the page on your device and never uploaded.
- 2
Paint over what should not be seen
Pick pixelate, blur or black out, then brush across the face, plate or text. It is hidden as you draw. The restore brush paints the original back if you go over the line.
- 3
Adjust, then save
Change the effect or its strength and every painted area updates at once. Undo works stroke by stroke. Download the result, or copy it to paste straight into a chat or a ticket.
Blur, pixelate or black out — which one hides it
Pixelate replaces each square of the picture with its average colour. It is the familiar look of a face hidden on the news, and at a high strength the squares are large enough that a face, a number plate or a house number simply is not there any more. The grid is fixed to the picture rather than to your brush, so neighbouring strokes line up and painting the same area twice never shifts it.
Blur smears each pixel across its neighbours. It reads as softer and less dramatic than pixelation — good for a crowd in the background of a photo, or a stranger’s face in a holiday picture you want to post. At low strength it is a privacy gesture rather than a guarantee; turn it up for anything that matters.
Black out paints the area solid black. It is the least pretty and the only one that removes information rather than disguising it, which is why it is the right choice for text.
Strength scales with the picture, so the same setting hides about the same share of a 12-megapixel phone photo as of a small screenshot. And because the brush only records where to hide, you can paint first and decide on the effect afterwards — every area changes together.
Why pixelated text is not really hidden
A face is unique, which is what makes pixelating one work: there is no way to guess what was under the squares. Text in a screenshot is the opposite. It is drawn in a known font, at a known size, from a small alphabet, so someone who wants to read it can type out candidate words, pixelate them the same way, and compare the resulting blocks against yours until they match. Tools that automate this are public, and they work best on exactly the thing people most often pixelate — a password, a name or an email address in a screenshot.
Blur has the same weakness, and a light blur is weaker still, since the shapes of the letters survive it. Black out does not: there is nothing left in the pixels to match against. So the rule of thumb is simple — blur or pixelate people and places, black out words.
The other leak is the file itself. A photo straight off a phone usually carries the coordinates where it was taken. The redacted copy is written fresh from its pixels, and that carries no metadata, so hiding a house number in the picture does not leave the house’s location in the file.
Frequently asked questions
No. The picture is decoded by your browser and every brush stroke is applied by code running on this page. There is no server step at all — you can load the page, disconnect from the internet, and redact and save as normal. For a tool whose whole purpose is hiding something, sending the unredacted original to a stranger's server first would defeat the point.
For faces and number plates in a photo, pixelate or blur at a high strength is fine and looks less heavy-handed. For text — names, email addresses, account numbers, anything typed — use black out. Blur and pixelation scramble information; black out removes it, and with text that difference matters.
Sometimes, yes. Text in a screenshot is drawn in a known font at a known size, so someone can render candidate words, pixelate them the same way and compare the blocks — public tools exist that do exactly this. Light blur has the same weakness. That is why black out is offered and why it is the right choice for text: there is nothing left to compare against.
Yes. Your strokes record where to hide, and the effect is applied on top of them. Switch from pixelate to black out, or turn the strength up, and every area you have already painted changes with it. Undo removes strokes one at a time, and redo puts them back.
Switch the brush to Restore and paint over the part that should show — the original comes back under the brush. Undo also works for whole strokes, and Clear removes everything at once.
No. The download is written fresh from the redacted pixels, and that process carries no EXIF, so the location, camera and timestamp a phone photo usually holds are not in the saved file. If you only want to remove metadata without changing the picture, the EXIF viewer does that losslessly.
The image is saved at its full original resolution. Screenshots and PNGs are saved as PNG, which is lossless, so text you did not redact stays crisp. Photos are saved as JPEG or WebP at high quality, which keeps the file a sensible size; a JPEG is re-encoded once in the process.
Yes — paint with a finger. On a small screen, zoom the brush down and work in short strokes; undo is one tap away if a stroke lands somewhere it should not.
Not with this tool — it works on images. For a PDF, take a screenshot of the page and redact that, or use the redaction feature of a PDF editor that actually deletes the underlying text, not one that only draws a box over it.
The Zeal Promise
Non-negotiable, on every tool we ship.
- No sign-up — ever
- No watermarks on any output
- No ads near the download button
- Works fully on mobile
- Privacy-first: processing stays in your browser
- Open source (MIT) — audit it, contribute, self-host
- Zero runtime dependencies in tool logic