Help

Frequently asked questions

Everything you might want to know before (or after) removing your first background.

Privacy

Are my images uploaded to a server?+

No. PNG Valley runs the AI model inside your browser using WebAssembly or WebGPU. Your photo is decoded, processed and exported on your own device. The only thing downloaded is the AI model itself, the first time you use the tool.

Do you store, see or train on my photos?+

We can't — the images never reach us. Nothing is sent for storage or training; everything stays in your browser tab.

Can I use the results commercially?+

The images you edit are yours, and we claim no rights over them. Just make sure you have the rights to the original photo you're editing. See our Terms of Use for details.

Using the tool

Is the background remover really free?+

Yes. There's no sign-up, no credits, no watermark and no resolution cap. Because the processing happens on your device, it doesn't cost us server time per image, which is what makes it sustainable to offer for free.

How do I remove the background from an image?+

Open the background remover, then drop an image onto the page, paste it with Ctrl/Cmd + V, browse for a file, or paste an image link. The background is removed automatically. You can refine edges with the brushes, pick a new background, and download a PNG, JPG or WebP.

Why does the first image take longer?+

The first time, your browser downloads the AI model (roughly 40–180 MB depending on your device and which engine runs). After that it's cached, so later images start almost instantly — even offline.

Can I remove backgrounds from many images at once?+

Yes. Batch mode processes up to 50 images in a queue, applies the same background and size to all of them, and lets you download everything as a single ZIP file.

Quality

What kinds of photos work best?+

Photos with a clear main subject — people, products, animals, cars, logos — work best. Good lighting and some contrast between subject and background help the AI find clean edges. Very busy scenes or subjects that blend into the background may need a quick touch-up with the restore or erase brush.

How do I fix hair, fur or missed areas?+

Open the Refine panel. Use the Restore brush to paint back parts that were removed and the Erase brush to remove leftovers. Lower the brush hardness for soft edges like hair. The Feather slider softens the whole edge and Edge shift grows or shrinks the cut-out by a few pixels.

Will the result keep my original resolution?+

Yes. Editing happens on a fast preview, but exports are rendered from your full-resolution original. You can also choose a custom size or a platform preset such as 2000×2000 for marketplaces.

Formats & export

Which file formats can I upload?+

JPG, PNG, WebP, HEIC/HEIF (iPhone photos, converted in your browser) and AVIF, up to 25 MB per image.

Which format should I download — PNG, JPG or WebP?+

Choose PNG for a transparent background (logos, design work, stickers). Choose JPG for a solid background when you need the widest compatibility, for example marketplace uploads. WebP keeps transparency with much smaller files, which is great for websites.

How do I make a white background for product photos?+

After the background is removed, pick white in the Background panel, or choose the “Amazon / Shopify product” preset in Crop. That preset centers your product on a pure white 2000×2000 canvas with even padding, ready to upload.

Can I add a shadow or outline to the cut-out?+

Yes. The Effects panel adds a soft drop shadow (offset, blur, opacity, color) and an outline stroke around the subject — useful for stickers, thumbnails and product shots. You can also adjust brightness, contrast and saturation.

Technical

Does it work on my phone?+

Yes, on modern iOS and Android browsers. Phones have less memory, so very large images may take longer.

Which browsers are supported?+

Recent versions of Chrome, Edge, Firefox and Safari. Browsers with WebGPU (such as recent Chrome and Edge) run the model on your graphics card for extra speed; others use WebAssembly on the CPU.

Which AI model does it use?+

The primary engine is an ISNet-based segmentation model packaged by IMG.LY. If it can't run on your device, the tool automatically falls back to the RMBG-1.4 model via transformers.js.

Looking for free PNG images and backgrounds? Browse PNG Valley.