Name: ezflows/Seedream 5.0 Pro Layer Decomposition
Price: 0.190629 USD

Call it

curl -X POST https://api.ezflows.io/v1/models/seedream-5-pro-layer-decomposition \
  -H "Authorization: Bearer $EZFLOWS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"image": "a wide shot of a desert at dusk"}'

# { "request_id": "req_8f2a41c9", "poll_url": "…", "status": "queued" }

Parameters

Name	Type	Default	Description
image required	string	—	The image to decompose. Exactly one image — png, jpeg, bmp, tiff or gif, up to 30 MB, between 512x512 and 6000x6000 pixels. Mark the elements you want separated: box one to name it and send its coordinates in the prompt, or point, circle and scribble to mark it on the picture alone.
prompt	string	—	Optional. List the elements to separate, one by one, naming each by its object and a distinguishing feature (colour, position). Leave empty and the model detects every major element itself. Each element you name becomes a billed layer, so listing them keeps the cost predictable. Boxing an element on the image adds its name and coordinates here as a <bbox> tag; the canvas rewrites those whenever the boxes change, so edit your own wording freely and leave the tags to it.
size	enum	auto	Resolution of the output base image and layers. auto follows the input image's own size; 1K, 1.5K and 2K pin it. Larger outputs cost more per image.
output_format	enum	png	File format of the base image. Every element layer is always PNG with a transparent channel, whatever this is set to.
optimize_prompt_mode	enum	standard	standard gives higher quality separation and takes longer; fast trades some quality for speed.

What a run costs

$0.191average · per run

$0.00for a failed run

~88smeasured mean, end to end

Compare every price

How it stacks up

Model	Price	Unit	Typical run
Seedream 5.0 Pro Layer Decomposition	avg $0.191	run	~88s
GPT Image 1.5	avg $0.18	run	~35s
GPT Image 1 Edit	avg $0.212	run	~43s
GPT Image 1	avg $0.217	run	~39s
Nano Banana Pro	avg $0.157	run	~42s