tencent/hunyuan-video

A state-of-the-art text-to-video generation model capable of creating high-quality videos with realistic motion from text descriptions

Capabilities

1:1 4:3 3:4 16:9 9:16 2:3 3:2 Seed

Community model (estimated from hardware time)

Name	Type	Description	Default	Constraints
`embedded_guidance_scale`	number	Guidance scale	`6`	min: 1, max: 10
`fps`	integer	Frames per second of the output video	`24`	min: 1
`height`	integer	Height of the video in pixels (must be divisible by 16)	`480`	min: 16, max: 1280
`infer_steps`	integer	Number of denoising steps	`50`	min: 1
`prompt`	string	The prompt to guide the video generation	`"A cat walks on the grass, realistic style"`	—
`seed`	integer	Random seed (leave empty for random)	`—`	—
`video_length`	integer	Number of frames to generate (must be 4k+1, ex: 49 or 129)	`129`	min: 1, max: 200
`width`	integer	Width of the video in pixels (must be divisible by 16)	`864`	min: 16, max: 1280

embedded_guidance_scale number

Guidance scale

Default: 6 min: 1, max: 10

fps integer

Frames per second of the output video

Default: 24 min: 1

height integer

Height of the video in pixels (must be divisible by 16)

Default: 480 min: 16, max: 1280

infer_steps integer

Number of denoising steps

Default: 50 min: 1

prompt string

The prompt to guide the video generation

Default: "A cat walks on the grass, realistic style"

seed integer

Random seed (leave empty for random)

video_length integer

Number of frames to generate (must be 4k+1, ex: 49 or 129)

Default: 129 min: 1, max: 200

width integer

Width of the video in pixels (must be divisible by 16)

Default: 864 min: 16, max: 1280

Version: 6c9132aee144 Updated: 2/26/2026 116.9K runs