Clean alpha edges
Hair, mesh, glass and soft shadows come out as proper transparency, not a hard halo.
TOOL · REMOVE BACKGROUNDAFFOGATO TOOLS
Subject in, background gone.
Remove Background cuts the subject out of an image and returns it on a transparent alpha channel. Upload, click, done — the model handles hair, fabric edges and fine detail so you don't have to mask anything by hand.
ONE CLICK · TRANSPARENT PNG · NO MASKING · NO WATERMARK
01WHAT YOU GET
Built for the job, not a generic prompt box.
No settings to learn
There's nothing to tune: upload the image, click Remove Background, download the PNG.
Built for product work
Cut-outs drop straight into marketplace listings, ads and the Product Studio for new scenes.
Chain it
Send any generated image to Remove Background, then on to Upscale or Clothes Swap without leaving the studio.
Current best model
Runs on a segmentation model (BiRefNet by default) that's upgraded as better ones land — no plugin updates on your side.
Commercial use included
Every plan includes a commercial license, and output has no watermark.
02HOW IT WORKS
How to remove an image background with AI, in three steps.
- 01
Upload the image
PNG, JPEG or WebP up to 20 MB — a product shot, a portrait, a generated image, anything with a clear subject.
- 02
Click Remove Background
There are no options. The model isolates the subject and builds the alpha channel.
- 03
Download or continue
Grab the transparent PNG from your assets, or send it to Upscale, Clothes Swap or the Product Studio for a new scene.
INPUT PNG · JPEG · WEBP ≤ 20 MBOUTPUT TRANSPARENT PNG (ALPHA)SETTINGS NONEMODEL BIREFNET (DEFAULT)
03EXPLAINED
How does an AI background remover work?
An AI background remover uses a segmentation model to decide, pixel by pixel, what is subject and what is background, then writes the result as an alpha channel — so the subject sits on true transparency you can place on any colour, scene or layout. Good models resolve hair, fabric edges and semi-transparent areas that used to need a manual mask.
Affogato's Remove Background is a single-click tool inside the Image Studio. It runs on a current segmentation model (BiRefNet by default, upgraded as better ones arrive) and outputs a PNG with a clean alpha channel.
It's most useful as a step in a larger flow: cut out a product, then build a new scene around it in the Product Studio; cut out a person, then swap the outfit with Clothes Swap; cut out a generated subject and upscale it for print.
04USE CASES
What people make with it.
Marketplace listings
Clean white-background cut-outs for Amazon, Shopify and app stores.
Ad and thumbnail assets
Transparent subjects that drop onto any layout, colour or headline.
Product scene building
Cut out the product once, then generate lifestyle scenes around it in the Product Studio.
Compositing people
Isolate a person for a new background, a poster or a group composite.
Stickers and icons
Turn generated art into transparent stickers and app assets.
05FAQ
Questions, answered.
What does the output look like?
A PNG of the subject on a transparent (alpha) background, at the input's resolution. Place it on any background in your design tool or in another Affogato tool.
Does it handle hair and fine edges?
Yes. The segmentation model produces soft, accurate alpha for hair, mesh, glass and shadows rather than a hard cut-out edge.
Are there settings?
No. Upload and click. If you need a specific new background, send the cut-out to the Image Editor or Product Studio and describe the scene.
How much does it cost?
It's priced per image in credits and the exact cost is shown before you run it. Plans start at $24/month with 2,000 credits that roll over.
Can I batch images?
Run it on as many images as you like, one click each, and download them together from your assets.
Can I use the cut-outs commercially?
Yes — every plan includes a commercial license with no watermarks.
Upload.
Click. Done.
Sign in, drop an image on Remove Background and download the transparent PNG.
NO WATERMARKS · CREDITS ROLL OVER · COMMERCIAL LICENSE INCLUDED · CANCEL ANYTIME