Skip to content

生数

Generate 16-second 1080P video with a U-ViT model

Visit official site

What 生数 is

The model is described as Vidu, a long-duration, consistent and highly dynamic video model released together with Tsinghua University, built on an original U-ViT architecture that fuses Diffusion and Transformer.

What you can do with it

  • Generate a 16-second video at up to 1080P with one click
  • Simulate the real physical world with rich imagination
  • Produce multi-shot video with varied viewpoints
  • Keep spatiotemporal consistency across the video

Application scenarios

  • Make a film trailer quickly
  • Generate simulated science experiment videos for education
  • Produce product introduction videos for e-commerce platforms
  • Simulate physical experiment processes in research

Who it is for

  • Companies and individuals needing high-definition video content
  • Professionals developing creative video content
  • Education work that needs teaching videos

Getting started

  • Visit the official site or platform of the model
  • Choose the video duration and resolution as needed
  • Enter or upload text descriptions, images or video material
  • Confirm the spatiotemporal consistency requirements of the video

Pros & cons

✓ What we like

  • One-click generation of 16-second 1080P video is listed
  • The architecture is described as supporting large-scale scalability checks

! What to watch out for

  • Generation is described through the U-ViT architecture
  • Video length is stated as up to 16 seconds

FAQ

How long can a video be?

The material states one-click generation of videos up to 16 seconds at 1080P.

What architecture is used?

An original U-ViT architecture fusing Diffusion and Transformer is described.

Last reviewed: 2026-11-15

More AI video generation tools

View all →

How we review