Cutout — Free AI Background Remover, 100% Client-Side. A privacy-first remove.bg alternative by @nap.codes.

By @nap.codes
READY v1.0 · onnx · in-browser AI
FILES 0 queued
+
Drop images here
or click to browse — multiple files supported
JPG · PNG · WEBP
Processing 0 of 0

100% client-side · no server · no uploads · Segmentation via @imgly/background-removal · Built by @nap.codes

A free remove.bg alternative that never uploads

Cutout runs an AI segmentation model directly in your browser — remove backgrounds from unlimited photos without giving them to anyone.

If you have been using remove.bg, Photoroom, or Adobe Express and hitting credit caps, watermarks on high-res exports, or "sign in to continue" walls, Cutout gives you the same neural background removal without any of that. The ONNX model runs on your device via WebAssembly. Nothing touches a server.

How Cutout compares

Featureremove.bgPhotoroomAdobe ExpressCutout
Free tier resolutionLow-res previewWatermarkedLimitedFull resolution
Batch / bulk processingPaid onlyPaid onlyNoUnlimited
Credits / signupCredits + accountAccountAdobe IDNone
ProcessingServer-sideServer-sideServer-sideClient-side
Images upload to a server?YesYesYesNo
Works offlineNoNoNoYes (after first load)
CostCreditsSubscriptionSubscriptionFree, always

Comparison based on publicly published features as of 2026. remove.bg, Photoroom, and Adobe Express are trademarks of their respective owners; this comparison is for informational purposes only.

Why in-browser matters

Every server-based background remover uploads your image to a data center in another country, runs a segmentation model on their GPU, and sends the cutout back. Your original photo now lives — however briefly — in someone else's storage bucket, subject to their retention policy and whatever regulation applies where their servers sit.

Cutout ships the model to your browser instead. The first run downloads about 44 MB of neural weights (a one-time cost, cached by the browser forever). Every image after that is processed locally using WebAssembly. Zero uploads. The Network tab will confirm it.

Built for these workflows

e
E-commerce catalogs
Product photos on white or transparent backgrounds. Batch-remove 100 SKU shots in one drop, then swap to a brand color background.
p
Portraits & content
Headshots for LinkedIn, Slack, or a team page. Portraits for thumbnails and social clips. Keep the subject, drop the room.
d
Design assets
Isolate props, objects, and stock elements to layer inside Figma, Canva, or InDesign. Transparent PNG, ready to drag.

Frequently asked

Is Cutout really free? What is the catch?

Yes — no signup, no daily cap, no watermark, no paid tier. There is no catch because there is no server cost to recoup: the model runs in your browser.

How is this different from remove.bg?

remove.bg uploads your image to their servers, runs a segmentation model there, and returns the result. Free users get preview-resolution output or must spend credits for full-resolution downloads. Cutout runs a comparable model in your browser using ONNX Runtime — full resolution, unlimited use, no credits, no uploads. The trade-off is the first-run model download (~44 MB, cached by the browser after).

Are my images really not uploaded?

Correct. Open your browser's Network tab while processing. You will see the initial model asset download (once), then zero image uploads. Photos stay on your device.

Which model tier should I pick?

Fast (~10 MB) for quick draft cutouts and previews. Balanced (~44 MB) for everyday use — the sweet spot for product photos and portraits. Accurate (~176 MB) when you need clean edges around hair, fur, or fine detail. Bigger models are slower per image but produce cleaner masks.

Why does the first image take longer than the rest?

The first run downloads the model weights and initializes the ONNX runtime in WebAssembly. After that, subsequent images process in a few seconds each on a modern desktop, longer on mobile depending on image size and device power.

Can I remove backgrounds in bulk?

Yes — drop as many images as your device can handle. Practical ceiling: 100-200 images on desktop, 20-40 on mobile depending on resolution. There is no server-imposed cap.

How accurate is it compared to paid tools?

For product photos, e-commerce shots, portraits with clean separation, and everyday objects, quality is very close to remove.bg. Wispy hair, semi-transparent glass, and thin lines can still be tricky — the same is true of every segmentation model. Use the Accurate tier for edge cases.

Does it work offline?

Yes, once the model is cached. Disconnect from the internet and keep processing images. Save the page (Ctrl+S / Cmd+S) for a fully local copy.

Can I use Cutout for commercial work?

Yes. Free for any use — personal, agency, e-commerce, freelance. No attribution required.

remove.bg® is a trademark of Kaleido AI GmbH. Photoroom is a trademark of Photoroom SAS. Adobe Express is a trademark of Adobe Inc. Cutout is an independent tool built by @nap.codes and is not affiliated with, endorsed by, or sponsored by any of these companies. Comparisons are based on publicly documented capabilities and are provided for informational purposes only.