Captured photo
text2video-zero
560

Percs

High quality
9:16 vertical

About

Text2Video-Zero leverages pre-trained text-to-image diffusion models to create video sequences from text descriptions. The model works zero-shot, meaning it requires no video-specific fine-tuning, instead using latent code interpolation and cross-frame attention mechanisms. Ideal for rapid prototyping of video concepts from written prompts.

Settings

Negative Prompt-  Negative Prompt
Video Length-  Video length in seconds
FPS-  video frames per second
Motion X-  Strength of the horizontal motion field.
Motion Y-  Strength of the vertical motion field.
Seed-  Random seed. Leave blank to randomize the seed