Skip to content

Upscale-A-Video

Raise video resolution with diffusion and a text prompt

Visit official site

What Upscale-A-Video is

The model is described as a diffusion-based approach that raises video resolution from a low-resolution video and a text prompt, using local and global mechanisms for time consistency.

What you can do with it

  • Raise the resolution of a low-resolution video with a text prompt
  • Keep short sequences consistent with temporal layers in U-Net
  • Use recurrent latent propagation for consistency across segments

Application scenarios

  • Improve old film clips
  • Raise the resolution of real-world video
  • Improve the visual quality of animated video

Pros & cons

✓ What we like

  • Text prompts can guide texture creation
  • A fine-tuned VAE-Decoder reduces remaining flicker

! What to watch out for

  • Both a low-resolution video and a text prompt are inputs
  • Restoration and generation are balanced by an adjustable noise level

FAQ

What are the inputs?

The material names a low-resolution video and a text prompt.

How is flicker handled?

A fine-tuned VAE-Decoder is described as reducing remaining flicker artifacts.

Last reviewed: 2026-11-15

More AI video generation tools

View all →

How we review