Background Remover
Cutting a subject out of its background by hand is slow work. This tool does the separation automatically, returning the subject on a transparent background so you can drop it onto a product page, a slide or a different photo.
It works best on images with a clear subject (a person, a product, an animal) against a background that is reasonably distinct from it.
Runs in your browser. Background Remover processes your input on your own device. Nothing you enter or upload is sent to a server.
Did this tool help you? ☕
WebzTools is free to use. If this tool saved you some time, you can support the project with a coffee and help us keep improving the tools.
(opens in a new tab)How to use the Background Remover
- Upload the image you want to cut out.
- Click Remove Background. The first time you do this, your browser downloads a small ML model; after that it's cached.
- Review the result against the original.
- Download the image with its transparent background. PNG is the default; WebP is an optional smaller choice.
Supported formats: Common web image formats; the result is a PNG by default, or a WebP if you choose it.
Things worth knowing
- The first use downloads a machine-learning model to your browser, which needs a working WebGPU or WebAssembly runtime and takes a moment on a slow connection.
- Fine edges such as hair, fur and transparent material are the most likely to need manual cleanup.
Frequently asked questions
Is my image uploaded to a server?
No. Background removal runs entirely in your browser using a local machine-learning model: the image, the generated mask and the result are never sent to a server.
What kind of images work best?
Images where the subject stands out clearly from what is behind it. Good lighting, a sharp subject and a background that differs in colour or focus all help. Fine detail such as loose hair, fur and semi-transparent fabric is the hardest case and may need manual touching up afterwards.
What format do I get back?
A PNG by default, because PNG supports transparency and keeps every pixel exactly. The PNG is losslessly optimized in your browser before it downloads, and the page shows the real file size and pixel dimensions first. You can instead choose WebP, which is usually much smaller and also keeps transparency, but compresses the colours, so they can shift slightly. A JPEG is not offered because it would fill the transparent area with a solid colour.
Why is my result bigger than the image I uploaded?
A transparent PNG stores every pixel, including the soft semi-transparent edges of the cutout, so it can be larger than a JPG or WebP original, which are compressed more aggressively. The pixel dimensions stay the same as your upload. If you need a smaller file, choose WebP before downloading, or resize the cutout with the Image Resizer, which changes pixel dimensions and keeps transparency.
Why does the first run take longer?
The first time you use this tool, your browser downloads the segmentation model (a few tens of megabytes) and caches it. Every run after that reuses the cached model and starts processing immediately.