ViduReferenceVideoNode - ComfyUI Built-in Node Documentation

The Vidu Reference Video Node generates videos from multiple reference images and a text prompt. It uses AI models to create consistent video content based on the provided images and description. The node supports various video settings including duration, aspect ratio, resolution, and movement control.

Inputs

Parameter	Description	Data Type	Required	Range
`model`	Model name for video generation (default: “viduq1”)	COMBO	Yes	`"viduq1"`
`images`	Images to use as references to generate a video with consistent subjects (maximum 7 images)	IMAGE	Yes	-
`prompt`	A textual description for video generation	STRING	Yes	-
`duration`	Duration of the output video in seconds (default: 5)	INT	No	5-5
`seed`	Seed for video generation (0 for random) (default: 0)	INT	No	0-2147483647
`aspect_ratio`	The aspect ratio of the output video	COMBO	No	`"16:9"` `"9:16"` `"1:1"`
`resolution`	Supported values may vary by model & duration (default: “1080p”)	COMBO	No	`"1080p"`
`movement_amplitude`	The movement amplitude of objects in the frame (default: “auto”)	COMBO	No	`"auto"` `"small"` `"medium"` `"large"`

Constraints and Limitations:

The prompt field is required and cannot be empty
Maximum of 7 images allowed for reference
Each image must have an aspect ratio between 1:4 and 4:1
Each image must have minimum dimensions of 128x128 pixels
Duration is fixed at 5 seconds

Outputs

Output Name	Description	Data Type
`output`	The generated video based on the reference images and prompt	VIDEO

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): ac86287d93257105446a5af42efd7d7fc07738fb7fcd0b260b8049a98552f193

​Inputs

​Outputs

Inputs

Outputs