Skip to content
BytePlusBytePlusBytePlus ModelArk for ComfyUI

Product Lookbook

Template
Product Lookbook, in the ComfyUI template browser under BytePlus
Workflow file
Product Lookbook.json
BytePlus nodes
Seedream 4.5 & 5.0
Keys needed
ModelArk API key
Also uses
Load Image, MarkdownNote, Save Image (built into ComfyUI)

What it makes

Section titled “What it makes”

Four lifestyle photos of the same product in four different places, from one request. It is for shops that need a set of consistent pictures for a product page or a catalogue. The idea: Seedream can return several related images from one prompt, so you describe all the scenes together, number them, and attach one reference photo to keep the product the same.

The workflow

Section titled “The workflow”
The Product Lookbook template: Load Image into the Seedream node, then Save Image
Two groups: the product photo, and the Seedream node with its Save Image.
  1. Load Image (titled “Product photo (change me)”) supplies the product photo. It points to a placeholder file, product_photo.png, so pick your own image.
  2. Seedream 4.5 & 5.0 (seedream 5.0 lite, titled “Sequential lookbook”) receives the photo as image_1. The prompt asks for a four-image lookbook with numbered scenes: a home office desk, a backpack on a hiking trail, a cafe table at golden hour and a gym bench. The size is 2304x1728 (4:3) and max_images is 4.
  3. Save Image (titled “Save the lookbook”) writes all four pictures to BytePlus/Image/lookbook.

Run it

Section titled “Run it”
  1. Add your ModelArk API key in Settings → BytePlus if you have not yet (API keys). This template needs no other key.
  2. Open the template browser and choose Product Lookbook under BytePlus.
  3. In Product photo (change me) choose your own image, then rewrite the prompt: describe your product in words and replace the four numbered scenes.
  4. Press Run. The four pictures arrive together in Save the lookbook.

Settings worth changing

Section titled “Settings worth changing”
  • Prompt: keep the pattern “same product in every image” plus a numbered list of scenes, and say what must stay the same, such as the shape and the cap.
  • max_images: match it to the number of scenes you wrote. The template note says up to 15 in total, with reference images counted.
  • size_preset: choose a ratio for where the pictures will be used, such as 1:1 or 3:4.
  • model: max_images exists on seedream 5.0 lite and on the 4.5 and 4.0 models, but not on 5.0 Pro or Flash. Keep one of those three for a sequential lookbook.
  • fail_on_partial: off here, so you still get the images that succeeded if one fails.

Example result

Section titled “Example result”
Four generated photos of a black bottle with a bamboo cap: on a desk, in a backpack pocket on a trail, on a cafe table and on a gym bench
Example result: four scenes, the same bottle in each.

Time and cost

Section titled “Time and cost”

The node bills on your ModelArk account for the images it generates. The seed control is set to randomize, so every run makes four new images and bills again. Set it to fixed and ComfyUI reuses the cached result of an unchanged workflow without billing.

Section titled “Related”