BriaProductHolding - ComfyUI Built-in Node Documentation - ComfyUI

This node puts a product in a person’s hands with Bria. The person’s identity, pose and background are kept while the scene is re-rendered at about 1 megapixel, so the result is not pixel-aligned with the input photo. All the product images describe one product from different angles or stages.

Inputs

The moderation selector reveals an additional moderation option when set to "true".

Common Inputs

Parameter	Description	Data Type	Required	Range
`person_image`	The photo of the person who holds the product. The alpha channel is dropped before the image is uploaded.	IMAGE	Yes	-
`product_images`	1 to 3 images of one product: the product itself, its packaging, another angle or a companion item. A batched input counts once per image and transparent areas are treated as white.	AUTOGROW	Yes	1 to 3 images
`instruction`	Optional extra direction added to Bria’s built-in instruction, such as which hand to use or which way the label faces. Default: `""` (empty string).	STRING	Yes	-
`aspect_ratio`	`"auto"` keeps the framing of the person photo at about 1 megapixel. A fixed ratio returns 0.6 to 1 megapixel, for example 768x1024 for `"3:4"`. Default: `"auto"`.	COMBO	Yes	`"auto"` `"1:1"` `"2:3"` `"3:2"` `"3:4"` `"4:3"` `"4:5"` `"5:4"` `"9:16"` `"16:9"`
`seed`	Seed sent to Bria with the request. Default: 42.	INT	Yes	1 to 2147483647
`moderation`	Moderation settings. When set to `"true"`, the moderation flag below becomes available.	DYNAMIC_COMBO	Yes	`"false"` `"true"`

Moderation “true” Inputs

This parameter appears only when moderation is set to "true".

Parameter	Description	Data Type	Required	Range
`visual_output_moderation`	Enables content moderation on the generated image. Default: `False`.	BOOLEAN	No	True / False

Constraints:

product_images accepts a maximum of 3 images, counted per image when the input is batched.
Transparent pixels in the product images are composited onto white before upload.
The alpha channel of person_image is dropped before upload.

Outputs

Output Name	Description	Data Type
`image`	The edited photo with the product placed in the person’s hands.	IMAGE

This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHub

Source fingerprint (SHA-256): 91a4a4dde0e9d7fe2480130da87edf1f8ed9d4ae0c86f8a8d030c36987dcdb8e