Free tool · Private · Runs in your browser

Image cleaner.

Remove the background from a photo, or strip the hidden data it carries: GPS location, camera, dates and more. Your image never leaves your device.

  1. 01

    Nothing is uploaded

    Every step runs on your device, in this page.

  2. 02

    Remove hidden data

    GPS location, camera, serial number, dates, editing history and comments.

  3. 03

    Remove the background

    An AI model cuts out the subject and gives you a transparent PNG.

Drop an image here

JPG, PNG or WebP. You can also paste an image with Ctrl+V.

Good to know.

Is my image uploaded anywhere?
No. Everything happens in your browser. The page downloads its code, and for background removal an AI model, but your image is never sent anywhere.
What is hidden data?
Photos carry more than the picture: the GPS location where they were taken, the date and time, the phone or camera model and its serial number, editing software, author names and sometimes a small preview copy. Removing it before you share a photo keeps that information private.
Does removing hidden data lower the quality?
No. For JPG, PNG and WebP the picture is copied exactly and only the extra data is cut out. Other formats, such as GIF and AVIF, are saved again as PNG.
Why keep the colour profile and orientation?
The colour profile tells screens how to show colours accurately, and the orientation keeps a phone photo the right way up. Neither holds personal information. Untick "Keep colour profile" to remove it too.
How does background removal work?
An AI model finds the main subject and makes everything else transparent. The first time, your browser downloads the model (44 MB for Fast, 88 MB for Best quality); after that it is cached and starts quickly.
Which files work?
JPG, PNG and WebP work best. iPhone HEIC photos only open in Safari; in other browsers, share or export them as JPG first.

Background removal uses the open-source IMG.LY background-removal library and model (AGPL-3.0), loaded from IMG.LY's servers. Photo details are read with exifr (MIT). The code of this page is on GitHub.