EmptyHunyuanImageLatent - ComfyUI Built-in Node Documentation

Complete documentation for the EmptyHunyuanImageLatent node in ComfyUI. Learn its inputs, outputs, parameters and usage.

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

The EmptyHunyuanImageLatent node creates an empty latent tensor with specific dimensions for use with Hunyuan image generation models. It generates a blank starting point that can be processed through subsequent nodes in the workflow. The node allows you to specify the width, height, and batch size of the latent space.

Inputs

Parameter	Data Type	Required	Range	Description
width	INT	Yes	64 to MAX_RESOLUTION	The width of the generated latent image in pixels (default: 2048, step: 32)
height	INT	Yes	64 to MAX_RESOLUTION	The height of the generated latent image in pixels (default: 2048, step: 32)
batch_size	INT	Yes	1 to 4096	The number of latent samples to generate in a batch (default: 1)

Outputs

Output Name	Data Type	Description
LATENT	LATENT	An empty latent tensor with the specified dimensions for Hunyuan image processing