zsxkib/pyramid-flow

Text-to-Video + Image-to-Video: Pyramid Flow Autoregressive Video Generation method based on Flow Matching

Capabilities

Reference Images

Community model (estimated from hardware time)

Name	Type	Description	Default	Constraints
`prompt` *	string	Text prompt for video generation	`—`	—
`duration`	integer	Duration of the video in seconds (1-3 for canonical mode, 1-10 for non-canonical mode)	`3`	min: 1, max: 10
`frames_per_second`	integer	Frames per second (8 or 24, only applicable in canonical mode)	`8`	8 24
`guidance_scale`	number	Guidance Scale for text-to-video generation	`9`	min: 1, max: 15
`image`	string (uri)	Optional input image for image-to-video generation	`—`	—
`video_guidance_scale`	number	Video Guidance Scale	`5`	min: 1, max: 15

prompt required string

Text prompt for video generation

duration integer

Duration of the video in seconds (1-3 for canonical mode, 1-10 for non-canonical mode)

Default: 3 min: 1, max: 10

frames_per_second integer

Frames per second (8 or 24, only applicable in canonical mode)

Default: 8

8 24

guidance_scale number

Guidance Scale for text-to-video generation

Default: 9 min: 1, max: 15

image string

Optional input image for image-to-video generation

video_guidance_scale number

Video Guidance Scale

Default: 5 min: 1, max: 15

Version: 8e221e66498a Updated: 6/8/2026 9.2K runs