Blur faces
Finds the faces in your photo and blurs them. Any face it misses, you draw over yourself. The photo is never uploaded — everything runs in this browser tab.
Face blur editor
Sample photos
0 blur areas. Tap an area to adjust it.
Why blur a face instead of covering it
Blurring mixes the colours inside the area you pick, so the shape of the face stops resolving into a face. A sticker or a solid block does something different: it leaves a hole that says a person stood here, and the expression and the direction they were looking go with it. When you want the photo to still read as the photo you took, blur is the gentler option.
The situations that bring people here are usually the same three. A club or company group shot is going somewhere public and one person in it never agreed to that. A photo of your own child is going outside the family and somebody else's child is in the frame. A photo of damage or an accident is going into a complaint or an insurance claim and a passer-by is in it. In all three the photo still has to work as a photo, so you cover the faces and nothing else.
What the blur strength number means
Strength is not a fixed number of pixels. It is a percentage of the short edge of the area you selected. A small face gets a proportionally smaller blur and a large face gets a larger one, so in a group photo somebody standing at the back ends up about as unreadable as somebody at the front.
| Blur strength | % of the short edge | On an area 200px across |
|---|---|---|
| 15 (weakest) | 5.9% | about 12px |
| 30 | 8.3% | about 17px |
| 60 (default) | 13.2% | about 26px |
| 100 (strongest) | 19.6% | about 39px |
Below about 30 the eyes and mouth still show faintly, and together with the rest of the photo that can be enough for someone to guess who it is. Blur at the default 60, then zoom in on the result before you save. If features are still coming through, raise the strength or enlarge the area.
Where people get caught out
Detection finds every face
Profiles are missed by design. The detector here is built for front-facing faces at a reasonable size. A head turned well to the side, a face that is only a few pixels across in a large group, or a face half cut off at the edge of the frame will be skipped. No setting fixes that — you tap the missed faces and place the areas yourself. Count the people in the photo before you save.
The detected box is big enough
Fitting the box tightly to the face leaves too much. An automatic box covers roughly the eyes, nose and mouth, which leaves the hair, the ears and the jawline outside it. Hair and ear shape are both things people recognise each other by. Pull the box out a little past the edge of the face.
Blurring erases the original
It covers, it does not erase. The file you download contains the blurred image, and we cannot promise a blur removes every trace of the detail underneath. Your original file is also still sitting on your device untouched, so check which file you are attaching when you send it. The one you download has -face-blur at the end of its name.
The saved file does not carry the original photo's metadata — no capture location, no camera or phone model. That is a side effect of how it is made: the image is redrawn onto a canvas in the browser and exported from there, and the original file's data does not come along.