Skip to content

FastVideo

Speed up large video diffusion models

Visit official site

What FastVideo is

The framework is described as an open source project for accelerating large video diffusion models through consistency distillation.

What you can do with it

  • Distill Hunyuan for faster video generation
  • Generate video quickly with FastMochi
  • Teach video diffusion in a course

What to watch out for

  • Development is described as ongoing and highly experimental
  • More distillation methods and models are listed as future plans
  • Setup starts from the provided environment script

Pros & cons

✓ What we like

  • Memory-efficient fine-tuning with LoRA is supported
  • Preprocessed data and pretrained weights are provided

! What to watch out for

  • Development is described as ongoing and highly experimental
  • More distillation methods and models are listed as future plans

FAQ

What speedup is stated?

An 8x inference speedup for the distilled models.

Which models are supported?

Mochi and Hunyuan video DiT models.

Last reviewed: 2026-11-15

More AI video generation tools

View all →

How we review