Wan2ReferenceVideoApi - ComfyUI Built-in Node Documentation - ComfyUI

This node generates a video featuring a person or object based on provided reference materials. It uses the Wan 2.7 model to create videos from a text prompt, supporting single-character performances and multi-character interactions. You must provide at least one reference video or reference image for the generation to work.

Inputs

Common Inputs

Parameter	Description	Data Type	Required	Range
`model`	The specific model to use for video generation.	DYNAMIC_COMBO	Yes	”wan2.7-r2v”
`seed`	Seed to use for generation, which helps control the randomness of the output (default: 0).	INT	Yes	0 to 2147483647
`watermark`	Whether to add an AI-generated watermark to the result (default: False). This is an advanced setting.	BOOLEAN	Yes	True False

wan2.7-r2v Inputs

Parameter	Description	Data Type	Required	Range
`prompt`	Prompt describing the video. Use identifiers such as ‘character1’ and ‘character2’ to refer to the reference characters. Must contain at least one character.	STRING	Yes	-
`negative_prompt`	Negative prompt describing what to avoid (default: empty).	STRING	No	-
`resolution`	The resolution of the output video.	COMBO	Yes	”720P" "1080P”
`ratio`	The aspect ratio of the output video.	COMBO	Yes	”16:9" "9:16" "1:1" "4:3" "3:4”
`duration`	The length of the generated video in seconds (default: 5).	INT	Yes	2 to 10

Reference Inputs

Parameter	Description	Data Type	Required	Range
`reference_videos`	Growable slot: connect up to 3 reference videos (slots `video1` to `video3`). At least one reference video or image is required overall.	VIDEO	No	0 to 3 items
`reference_images`	Growable slot: connect up to 5 reference images (slots `image1` to `image5`). At least one reference video or image is required overall.	IMAGE	No	0 to 5 items

Important Constraints:

You must provide at least one reference video or reference image in the reference_videos or reference_images inputs.
The total combined number of reference videos and reference images cannot exceed 5.
The prompt input must contain at least one character.

Outputs

Output Name	Description	Data Type
`output`	The generated video file.	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): 52ac550522bf3fe8f57444ce8586fe83be470b893ff8c01292743553cfbd623d