9 open-source video generation models to know in 2026, from Wan2.2 to HunyuanVideo 1.5
TheTuringPost · x · 2026-09-27
Turing Post rounds up nine open-source video generation model families worth knowing in 2026, spanning text-to-video, image-to-video, editing, audio-video generation, and local workflows.
Highlights:
- Wan2.2 (Alibaba): Apache 2.0 flagship with a 27B Mixture-of-Experts design (separate high-noise/low-noise experts) plus a dense TI2V-5B model that runs 720p/24fps text- and image-to-video on a single consumer GPU via a high-compression VAE.
- Wan2.1: the low-VRAM entry point — the 1.3B model needs 8.19GB VRAM for 5-second 480p clips; suite covers editing, text-to-image, and video-to-audio.
- HunyuanVideo 1.5 (Tencent): November 2025 rebuild cut parameters from 13B to 8.3B while improving quality, and runs on consumer hardware.
- Also covered: LTX-2/LTX-2.5, MiniMax H3, CogVideoX, Mochi 1, Kandinsky 5.0, Stable Video Diffusion.
The models differ sharply in quality, speed, hardware requirements, and license terms; the guide helps readers pick by use case.
Related event: Open-Source Video Generation Models to Watch in 2026(3 posts)→
More from Multimodal
- Gemini Live Avatars go GA: real-time talking avatars in 97 languages, with pricing — Sam Witteveen · 2026-09-27
- Opus 5.5 animates Pink Floyd's "Time" from a style ref and a few prompts — Hesamation · 2026-09-27
- 9 open-source video generation model families to know in 2026, from Wan2.2 to Mochi 1 — TheTuringPost · 2026-09-27
- Astra 6 vs Opus 5.5 for motion graphics: one takes 6 tries, the other 46 minutes — aziz4ai · 2026-09-27
- Shared prompt template: 'Wild Magic Transformation' image generation with vines, dragon scales and elemental flames — LudovicCreator · 2026-09-27
- Veda Sparse ships 10% keep-ratio block-sparse attention accelerator for MiniMax H3 T2VA — linoy_tsaban · 2026-09-27