gen‑ai.news
Open weightsFree

FramePack

Stanford (Lvmin Zhang)

FramePack is a video generation model line from Stanford (Lvmin Zhang). The latest release is FramePack (18 Apr 2025). Specs, licensing, and availability below. Official site →

FramePack

Open weightsFreeCommunity APISelf-host

Constant-memory next-frame-section prediction; 60s video on 6GB VRAM; F1 anti-drift variant followed

License
Apache 2.0
Commercial use
Allowed
Availability
Self-host, Community API

See FramePack in context on the video model timeline, or browse all model families.