BLUR · PIXELATE · SOLID BOX · NO UPLOAD
Blur Faces and Private Details
Hide a face in a crowd, a car’s plate or a name on a screenshot before you share it. Draw over what should stay private; the saved picture contains only the blur.
Choose the photo or screenshot with something to hide.
Loading Image Studio…
- No upload
- No sign-up
- No watermark
- No daily limit
- Works offline
How to blur part of a photo
Open the photo
Drop the picture on the page, paste a screenshot, or tap Open image. It opens in the editor with the Redact tool ready in blur mode.
Draw over each face or detail
Drag a rectangle over everything that should be hidden, one region at a time, and set the strength. Each region appears in the Objects list, where you can adjust or delete it.
Export a clean copy
Open Export and save. The new file is drawn from scratch with the blur baked into the pixels, and the original’s metadata is not copied.
FEATURES
What this page does
Blur, pixelate or solid box
Three ways to hide a region. Blur looks natural on faces, pixelate gives the familiar mosaic, and a solid box removes the content completely.
Minimum strength built in
Blur and pixelate have a floor that can’t be lowered, so a region can’t be left so lightly blurred that it is still easy to read.
Always a new file
An image with a redaction is never saved by copying the original’s bytes, even if that would be smaller. The export is always freshly encoded from the hidden version.
Hidden data removed
Location, the embedded thumbnail and other metadata are removed, so a small preview of the unblurred picture can’t ride along inside the file.
Works by keyboard and touch
Add a region with a button and set its position and size in number fields, or draw it with a finger on a phone.
READ THIS FIRST
Blur can sometimes be undone. A solid box can’t.
Blur and pixelate scramble a region, but they keep a trace of what was there. For a face in a crowd that is fine: nobody can rebuild a face from a soft smudge. For short, predictable text it is not. Researchers have recovered blurred or pixelated numbers, names and passwords by trying candidates until one blurs to the same pattern.
So the rule is simple. Blur or pixelate faces, bodies, backgrounds and anything where the shape alone gives nothing away. Use Solid box for passwords, email addresses, phone and card numbers, account IDs, addresses and anything else made of letters and digits. The blur panel repeats this note, so it is in front of you at the moment you choose.
| What you are hiding | Use | Why |
|---|---|---|
| A face or person | Blur | Looks natural and removes the detail a face is recognized by |
| A license plate | Solid box or strong pixelate | Plates follow known patterns that a guess can match |
| Passwords, emails, numbers | Solid box | Short text can be reconstructed from blur |
| A screen or document in the background | Blur | Removes legibility while keeping the scene |
What the saved file no longer contains
Phone photos carry more than the picture. EXIF data can include the exact GPS location, the date and the camera, and many JPGs also hold a small embedded thumbnail. Some tools edit the main image but leave that thumbnail alone, so the unblurred picture is still inside the file in miniature.
Here a redacted image is always rendered again and encoded as a new file. Location, the embedded thumbnail and other metadata are removed, and that stays true even if you turn on Keep date and camera, which keeps only the date and camera model. The redaction is painted into the pixels themselves; there are no layers to peel back.
You choose every region
There is no automatic face detection on this page. That is a deliberate trade: a detector that misses one face in a group photo is worse than none, because it makes you trust it. You see the picture, you draw a region over each face or detail, and nothing is hidden or left visible without your decision.
For a group shot, zoom in, draw the first region, then duplicate it with Ctrl+D and drag the copy over the next face. Regions can overlap, and the Objects list names each one so you can check that none was missed before you export.
Hidden regions are not applied
You can hide a region in the Objects list to peek at what is under it. A hidden region is not part of the export, and the Export panel warns you how many hidden redactions there are before you save, so a peek can’t accidentally become the file you share.
Is it really private?
The photos people blur are by definition the sensitive ones: children, colleagues, documents, screens. Every step here, from opening the image to drawing the blur and encoding the result, runs inside your browser tab. The unblurred photo never leaves your device.
While you edit, the editor keeps your work in this browser’s private storage, where the browser allows it, so a reload doesn’t lose it, and Recent edits on the start screen can reopen it. That copy includes the original photo, so use Remove, or Remove all, there when you are done with it.
You don't have to take our word for it. Three checks:
- Open your browser's developer tools, choose the Network tab, then open and save an image. No request carries your file.
- Open an image here once, then turn on airplane mode and run the same job again. It still works, because nothing needs a server.
- Read the page's response headers: its Content-Security-Policy lets it connect only to this site itself, so there is no outside host a file could go to.
Reference
| Opens | JPG, PNG, WebP, AVIF, GIF (first frame), BMP, ICO, SVG and TIFF |
|---|---|
| Saves | The same format by default; JPG, PNG or WebP on request |
| Modes | Blur, pixelate and solid box, with a minimum strength for blur and pixelate |
| Export | Always a fresh encode; never the original file’s bytes |
| Metadata | Location, the embedded thumbnail and other metadata are removed |
| Keyboard | R opens Redact; arrow keys move the selected region, Delete removes it |
Frequently asked questions
Can a blurred face be unblurred?
A strong blur throws away the fine detail a face is recognized by, so a face in a group or crowd is very unlikely to be recovered. The minimum strength is measured against the whole picture, not the face: when a face fills much of the frame, raise the strength until it is unrecognizable in the preview, or use Solid box. Short text is different: blurred or pixelated words and numbers can sometimes be guessed back. Use Solid box for anything written.
Is pixelating safer than blurring?
Not really. Both can leak short text, and both are fine for faces at a strong enough setting. Pixelate is the better-known look; blur is less distracting. For secrets made of characters, only a solid box removes the information.
Does the file still contain the original picture?
No. The export is drawn and encoded from scratch with the redaction in the pixels, and the embedded EXIF thumbnail, which could show the unblurred photo, is removed with the rest of the metadata.
Can I blur a face automatically?
Not on this page. You place each region yourself, which takes a few seconds per face and means nothing is missed because a detector failed to spot it.
Can I blur the background instead of a face?
Draw a large region over the background and the subject stays sharp only where you leave it uncovered. Several regions can be combined to follow the shape of a scene.
How do I blur part of a screenshot?
Paste it with Ctrl+V (⌘V on a Mac), switch to Solid box for any text such as names, emails or numbers, and blur pictures and avatars. Then export as PNG to keep the rest of the screenshot sharp.
Can I undo a blur after saving?
In the saved file, no; that is the point. In the editor, yes: undo, delete the region or change its strength any time before you export, and keep your original file as it was.