BriaEraseByText - ComfyUI Built-in Node Documentation - ComfyUI

This node removes an object described in plain text from an image using Bria. Bria re-renders the whole frame at about 1 megapixel, so the result is not pixel-aligned with the input.

Inputs

Common Inputs

Parameter	Description	Data Type	Required	Range
`image`	The image from which the named object should be removed.	IMAGE	Yes	-
`object_name`	Name of the object to remove, such as ‘the lamp’. Several objects can be named at once, such as ‘the phone and the pencils’. Naming something that is not in the picture still returns a re-rendered image, and the request is still billed. Must be at least 1 character long (default: empty).	STRING	Yes	-
`moderation`	Moderation settings. Selects whether the optional moderation controls are shown.	DYNAMIC_COMBO	Yes	`"false"` `"true"`

Moderation Inputs

These parameters appear when moderation is set to "true".

Parameter	Description	Data Type	Required	Range
`visual_input_moderation`	Enables content moderation on the input image (default: false).	BOOLEAN	No	true false
`visual_output_moderation`	Enables content moderation on the generated output image (default: false).	BOOLEAN	No	true false

Outputs

Output Name	Description	Data Type
`IMAGE`	The re-rendered image with the named object removed.	IMAGE
`structured_prompt`	Structured description of the edited image, for a follow-up edit with Bria FIBO Image Edit.	STRING

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): 51ac362bea731251c99905172c41cdebb5165e7564c508f13aa43cf9072964ef