Runway.

Local generative media tools when a hosted video AI is the wrong fit.

9 listedCatalog Sep 1, 2026

Rank 1. Wan2.1Alibaba’s open text-to-video and image-to-video stack.Apache 2.0 · LinuxGitHub stars: 14k
Rank 2. Open-SoraApache-2.0 research stack for training and running video diffusion on 24 GB cards.Apache 2.0 · Linux, macOS, +1GitHub stars: 30k
Rank 3. CogVideoXTHUDM open video diffusion with strong prompt following.Apache 2.0 · LinuxGitHub stars: 13k
Rank 4. FramePackLong local image-to-video with a low VRAM floor.Apache 2.0 · Linux, WindowsGitHub stars: 18k
Rank 5. LTX-VideoLightricks open video model aimed at faster local clips.Apache 2.0 · Linux, WindowsGitHub stars: 8k
Rank 6. MochiGenmo’s Apache-2.0 10B video model — permissive weights, heavy hardware.Apache 2.0 · Linux, macOS, +1GitHub stars: 3.7k
Rank 7. SkyReels V1HunyuanVideo-based human-centric video model that fits one 24 GB GPU.Other · Linux, macOS, +1GitHub stars: 2.7k
Rank 8. ComfyUINode graph for diffusion workflows. The power-user default.GPL-3.0 · Linux, Windows, +1GitHub stars: 91k
Rank 9. HunyuanVideoTencent’s 13B video generation model — top-tier motion, heavy VRAM needs.Other · Linux, macOS, +1GitHub stars: 13k