Ollama for video models.
Run open video models on your own GPU — install · pull · run — plus a drop-in skill so any coding agent can generate high-quality video.
📦 open-video · 🌐 open-video.ai · ▶ demo
- Today (v0.0.1): local MiniMax H3 — one-line install, Ollama-style CLI, agent skill harness. Apache-2.0 code; model weights carry their own upstream license.
- Next: real vision-judge loop, multi-shot films, more open model backends.
Contributions welcome — start at the open-video repo.