EmptyMochiLatentVideo - ComfyUI Built-in Node Documentation - ComfyUI

EmptyMochiLatentVideo creates an empty latent video tensor with the dimensions you specify. It generates a zero-filled latent representation that can be used as a starting point for video generation workflows. The node lets you define the width, height, length, and batch size of the latent video tensor.

Inputs

Parameter	Description	Data Type	Required	Range
`width`	The width of the latent video in pixels (default: 848, must be divisible by 16)	INT	Yes	16 to MAX_RESOLUTION
`height`	The height of the latent video in pixels (default: 480, must be divisible by 16)	INT	Yes	16 to MAX_RESOLUTION
`length`	The number of frames in the latent video (default: 25, must satisfy that `(length - 1)` is divisible by 6)	INT	Yes	7 to MAX_RESOLUTION
`batch_size`	The number of latent videos to generate in a batch (default: 1)	INT	No	1 to 4096

Note: The actual latent dimensions are calculated as width/8 and height/8, the temporal dimension is calculated as ((length - 1) // 6) + 1, and the tensor has 12 channels. The length parameter must satisfy that (length - 1) is divisible by 6, meaning valid values are 7, 13, 19, 25, etc.

Outputs

Output Name	Description	Data Type
`samples`	An empty latent video tensor with the specified dimensions, containing all zeros	LATENT

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): 1774e1b54b429a946172ba9f609b433d99c0ca2ced2d9e0e3b0b85c82e5141b2