Blur, pixelate or black out part of an image
Draw over the faces, plates or text you want hidden, pick an effect, and save a copy that keeps everything else sharp.
- Processed on your device. Your photos are never uploaded. How we know
- Free, no sign-up, no watermark
Drop an image to blur or pixelate
Drop files, paste with Ctrl+V, or choose them from your device.
PixWrench’s code runs on your device. We don’t receive your files, file names or their contents.
PixWrench’s code runs on your device. We don’t receive your files, file names or their contents. What happens to your files
How to blur a face, number plate or text in a photo
Before you post a street scene, sell a car or share a screenshot, there is often something in the frame that shouldn’t travel with it: a passer-by’s face, a license plate, an address on a parcel, an email in the corner of a screen. This tool covers those areas and leaves the rest of the picture alone.
You pick every region yourself. There is no automatic face detection, so nothing is hidden or missed without you seeing it. Hidden details in the file, like GPS position, are a separate issue; the EXIF viewer and remover deals with those.
- 1
Open the image
Drop in one photo or screenshot. JPG, PNG, WebP, AVIF and HEIC are accepted.
- 2
Cover each area
Draw a rectangle or ellipse, or add one with the button and steer it with the arrow keys. Choose pixelate, blur or solid fill for each.
- 3
Check and save
Look through the region list, zoom in on the edges, then download the full-resolution result.
What this tool does
- Rectangle and ellipse regions, as many as you need
- Per-region effect: pixelate (block size), blur (radius) or solid fill
- Solid fill in black or any color you choose
- Add regions with a button and move or resize them with the arrow keys
- A list of every region so none gets forgotten
- Saved in the original format at full resolution
- Metadata removed by default
What pixelation does to a number plate
Say a plate covers a 240 × 80 pixel area of a 1600 × 1200 photo. That area holds 19,200 pixels. With a pixelate block size of 16 px, it becomes a grid of 15 × 5 = 75 blocks, each a single flat color. Double the block size to 32 px and only about 8 × 3 blocks remain (the edge blocks are partial), which is far harder to read.
A blur with a larger radius spreads each pixel over a wider neighborhood instead of chopping the area into blocks. A solid fill replaces all 19,200 pixels with one color, leaving nothing of the original behind. Only the pixels inside your regions change; the rest of the photo stays at its full 1600 × 1200 resolution.
What each effect does to the pixels
| Effect | What it changes | Best for |
|---|---|---|
| Pixelate | Splits the region into square blocks, each filled with one color | Faces and backgrounds where a softer look is fine |
| Blur | Mixes each pixel with its neighbors within the radius | Faces and distracting objects |
| Solid fill | Overwrites the region with black or a color you choose | Text, numbers, documents, anything that must be unreadable |
The edited image is rendered in your browser and saved in its original format at full size. Because the effect is applied to the pixels themselves, there is no hidden layer underneath that someone could switch off later.
Why blur isn’t enough for text
Blur and pixelation reduce information; they don’t always destroy it. In a 2016 paper, researchers Richard McPherson, Reza Shokri and Vitaly Shmatikov showed that machine-learning models could still recognize faces and handwritten digits after mosaicing and blurring had been applied. The open-source Depix project demonstrates recovering text from pixelated screenshots. Text is a particular risk because it is built from a small set of characters, which narrows the guessing.
That’s why the tool recommends solid fill for passwords, account numbers, addresses and names. A flat box leaves nothing to reverse.
Good to know
- One image at a time, and regions are drawn by hand. Check every face in a crowd photo.
- Draw regions a little larger than the thing you’re hiding; letters that poke out of the edge are still readable.
- If the sensitive part is near an edge, cropping it out with the crop tool removes it completely.
- Hiding a face doesn’t hide everything: clothing, tattoos or a reflection can still identify someone.
- Keep the unedited original somewhere private if you may need it, and share only the edited copy.
What happens to your files
Processed on your device. Your photos are never uploaded. PixWrench’s code runs on your device. We don’t receive your files, file names or their contents.
The photo is opened and edited inside this browser tab, and the covered version is produced on your device. The original, the regions you draw and the result stay with you. Metadata such as location is removed from the output by default; our guide to removing location from photos explains why that matters.
Ad and analytics scripts load on this page too; they run separately from the editor and are described in our privacy policy.
Related tools and guides
Questions people ask
What is the safest way to hide text in a screenshot?
Use solid fill. Blurred or pixelated text can sometimes be reconstructed; a flat box can’t.
Can someone unblur my photo?
The saved file contains only the blurred pixels, not the original. But strong models can sometimes guess what was there, especially for text, so use solid fill for anything critical.
Does the tool find faces automatically?
No. You choose every region, which means nothing is hidden or skipped without you knowing.
Will the rest of my photo lose quality?
It is saved at full resolution in the same format. Only the regions you drew are changed.
Can I add a watermark at the same time?
Not in this tool. Save the edited image, then add text or a logo with the watermark tool.