ブラウザで画像の背景をすぐに透過

This is an online background remover that runs entirely inside your web browser. Instead of sending your photo to a server, a background-removal model is downloaded once to your device and then executed locally using WebGPU where available, or WebAssembly on older hardware. That single architectural choice is what makes the rest of the promises possible: no account, no daily limits, no watermark, and no place for your image to be stored, logged or reviewed.

The tool accepts JPG, PNG, WebP and AVIF images and returns a transparent-background PNG at the full resolution of your original, or a flattened JPG or WebP if you would rather have a solid background.

Product photos are the most common use case. If you sell on a marketplace, handmade store or social shop, listing images usually need a clean, plain background. Upload a photo, let the model isolate the product, choose white, and download an image that is ready to attach to a listing — without a subscription and without a per-image credit.

Profile pictures are the second: LinkedIn, work chat, forums and email. A transparent cutout lets you place yourself on any colour, and a light grey or white background gives a tidy passport-style result. Logos and graphics also work well, because flat artwork has hard edges that segmentation models handle reliably. Finally, social media assets — thumbnails, story backgrounds, collages and stickers — all benefit from a cutout you can layer over anything.

Most online image tools work by uploading your file to a server, processing it there, and returning a result. That means your photo leaves your device, travels over the network, sits in temporary storage on hardware you do not control, and is subject to a retention policy you have to take on trust.

This tool does not work that way. Your image is decoded, processed and re-encoded inside the browser tab, and the only network traffic is the one-time download of the model and runtime files on your first visit. There is no upload request, because there is no server endpoint that accepts images. You do not have to trust a privacy policy: you can open your browser developer tools, watch the Network panel while removing a background, and confirm that no request contains your photo. Even better, the tool continues to work with the internet disconnected.

The interface is deliberately small. Add an image, compare before and after, choose a background, download. A compare control lets you switch between a side-by-side and a stacked layout and zoom in to inspect hair and edges before committing, because that is where background removal usually shows its flaws.

The interface is available in eleven languages — English, Spanish, Portuguese, German, French, Hindi, Japanese, Korean, Simplified Chinese, Italian and Russian — so you can work in the language you are most comfortable reading.

Background removal quality depends far more on the photo than on the tool. Even, diffused lighting produces cleaner edges than harsh shadows. A subject that contrasts with its background is easier to isolate than a white shirt against a white wall. Sharp focus helps, and small images with fine detail lose fidelity faster than larger ones. When a result is imperfect, check it at 100% zoom before you publish it, and remember that transparent objects and very fine hair remain genuinely difficult for every segmentation model.