Blog
Replicate
STABLE

Stable Video Diffusion

SVD is a research-only image to video model

Inputs
Input Image
Input image
Cond Aug
Amount of noise to add to input image
Decoding T
Number of frames to decode at a time
Frames Per Second
Frames per second
Motion Bucket Id
Increase overall motion in the generated video
Seed
Random seed. Leave blank to randomize the seed
sizing_strategy
Decide how to resize the input image
video_length
Use svd to generate 14 frames or svd_xt for 25 frames
Outputs
Urls
URLs of the videos