Flux2ImageNode - ComfyUI Built-in Node Documentation - ComfyUI

Flux.2 Image Generate images using the Flux.2 [pro] or Flux.2 [max] model from a text prompt and optional reference images. This node sends your request to the BFL API, polls for the result, and returns the generated image as a tensor.

Inputs

Common Inputs

Parameter	Description	Data Type	Required	Range
`model`	The Flux.2 model version to use. Selecting a model unlocks additional parameters for width, height, and optional reference images.	DYNAMIC_COMBO	Yes	”Flux.2 [pro]" "Flux.2 [max]“
`prompt`	Prompt for the image generation or edit (default: empty string).	STRING	Yes	N/A
`seed`	The random seed used for creating the noise. Can be set to randomize after each generation (default: 0).	INT	Yes	0 to 18446744073709551615

Flux.2 [pro] and Flux.2 [max] Inputs

Shared by both models — the parameter sets are identical.

Parameter	Description	Data Type	Required	Range
`model.width`	The width of the generated image in pixels (default: 1024).	INT	Yes	256 to 2048 (step 32)
`model.height`	The height of the generated image in pixels (default: 768).	INT	Yes	256 to 2048 (step 32)

Reference Inputs

Parameter	Description	Data Type	Required	Range
`model.images`	Optional reference image(s) for image-to-image generation. Up to 8 images. Growable slot: connect 1..8 items (`image_1`…`image_8`).	IMAGE	No	0 to 8 images

Note:

The maximum number of reference images is 8. If more than 8 images are provided, an error is raised.
The model.width and model.height values affect the generation cost. The cost also depends on the selected model and on whether reference images are provided.

Outputs

Output Name	Description	Data Type
`image`	The generated image as a tensor, downloaded from the BFL API result.	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): 2994564757e1c66ac6da7b45d227b27ceb0020ac6fc9e8cbe2b53fe9f70bc195