FastH3 V2
An open-weight video diffusion model released by the FastVideo team. Its key characteristic is the ability to complete video generation in only eight sampling steps. It supports text-to-video and image-to-video generation, integrates a ComfyUI workflow, and is designed for local deployment scenarios.
Timeline (last 90 days)
The FastVideo team published the open-weight model for FastH3 V2 on Hugging Face.
FastH3 V2 can complete video generation in as few as 8 sampling steps.
FastH3 V2 is the latest iteration of the video diffusion model released by FastVideo.
The model weights repository address for FastH3 V2 is FastVideo/FastVideo-FastH3-8-Step-V2.
All Facts (4)
FastH3 V2 can complete video generation in as few as 8 sampling steps.
50%UnverifiedFastH3 V2 is the latest iteration of the video diffusion model released by FastVideo.
50%UnverifiedThe model weights repository address for FastH3 V2 is FastVideo/FastVideo-FastH3-8-Step-V2.
50%UnverifiedThe FastVideo team published the open-weight model for FastH3 V2 on Hugging Face.
50%