Great Work Labs
Models / Vision

Cutout

Background removal
cutout-1

Transparent PNG of the subject, background gone.

Price$2.00 per 1,000 images
Speed (our servers)5062 to 5062 ms
LanguagesNot applicable
InputImage, up to 10 MB
Built onDIS ISNet general-use (Apache-2.0)

Cutout removes the background from a photo and returns a transparent PNG of the subject: products, people, pets, cars. Ask for format: "mask" to get the black-and-white mask instead and composite it yourself.

Use it

curl https://labs.greatwork.company/v1/run \
  -H "Authorization: Bearer $GWL_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "cutout-1",
  "input": "https://labs.greatwork.company/examples/product.jpg"
}'

Set GWL_API_KEY to a key from your dashboard. The bare name cutout works too and follows new versions.

Options

input: An image: an https URL, a data URL (`data:image/png;base64,...`), or a file in a multipart upload. JPEG, PNG, WebP, GIF, AVIF or TIFF, up to 10 MB.

NameTypeDefaultWhat it does
formatstring
png | mask
"png"png for the cutout with transparency, mask for the alpha mask.

Output

FieldTypeMeaning
imagestringPNG as a data URL.
foregroundnumberShare of the image kept, 0 to 1.
widthintegerOutput width (long side up to 2,048 px).
heightintegerOutput height.

Every response wraps this in {"id", "model", "output", "usage": {"units", "unit", "cost_usd"}, "latency_ms"}.

Examples

Product

{
  "model": "cutout-1",
  "input": "/examples/product.jpg"
}
{
  "image": "/examples/out/cutout-1-0.png",
  "format": "png",
  "width": 1024,
  "height": 1024,
  "foreground": 0.146
}

Captured on our production servers: 5062 ms, 1 image, $0.002.

Good for

  • Product photos on white or transparent
  • Profile pictures and stickers
  • Thumbnails with the subject cut out
  • Masks for image editing

Limits and caveats

  • Hair and fur edges are soft, not strand-perfect.
  • Busy scenes with several subjects may keep more than you want.
  • Takes a few seconds per image.

Pricing

$2.00 per 1,000 images. One unit per image. Failed calls aren't charged, and a call never costs more than its estimate (POST /v1/estimate tells you before you run it).

Built on

DIS ISNet general-use (xuebinqin/DIS), 44M parameters. License: Apache-2.0 (ONNX export by img.ly, MIT). We run it on our servers.