Wan2.2: Next-Generation MoE Video Generation by Alibaba Cloud
Wan2.2 is Alibaba Cloud's next-generation video generation model with MoE architecture. Available in 14B T2V/I2V and 5B hybrid variants with film-level aesthetic control.
Wan2.2
VideoText-to-VideoImage-to-VideoOpen SourceMoESecond-generation Wan video model — Alibaba Cloud's next-generation video generation model released July 2025. Innovative MoE (Mixture of Experts) architecture with high-noise and low-noise expert models. Film-level aesthetic control, large-scale complex motion, and precise semantic compliance. Apache 2.0 licensed.
| Developer | Alibaba Cloud (Wan AI) |
| Release Date | 2025-07-28 |
| Architecture | MoE (Mixture of Experts) — high-noise + low-noise experts |
| License | Apache-2.0 |
| Model Sizes | 14B (dedicated T2V/I2V) / 5B (hybrid TI2V) |
| Capabilities | Text-to-Video (14B), Image-to-Video (14B), Hybrid T2V+I2V (5B), First-Last Frame, S2V, Fun Control, Animate |
| Output Resolutions | Up to 720P |
Guides and workflows related to this model series.
Wan2.2 Workflows in ComfyUI: T2V, I2V and GGUF Video Guide
Run Wan2.2 (14B/5B) workflows in ComfyUI: T2V, I2V, V2V. Comfy-Org repackaged wan2.2_ti2v_5B_fp16 and 14B fp8 models, video_wan2_2_5B_ti2v.json template, VAE links.
Wan2.2 Animate ComfyUI Workflow
Animate characters and replace video subjects with Wan2.2 Animate in ComfyUI. Precise motion transfer, expression replication for AI-driven character animation workflows.
Wan2.2 Fun Control ComfyUI: Official, Kijai and GGUF Workflows
Complete Wan2.2 Fun Control video generation in ComfyUI: official native version, Kijai WanVideoWrapper and GGUF quantized workflows for I2V, V2V and feature-guided video control.
Wan2.2 Fun InP ComfyUI Workflow Complete Usage Guide, Official + Community Versions (Kijai, GGUF)
The most comprehensive Wan2.2 Fun InP video ComfyUI tutorial: including official native version, Kijai WanVideoWrapper, GGUF quantized version and other first-last frame video generation workflows
Wan2.2-S2V Audio-Driven Video in ComfyUI: Complete Guide
Create talking avatars with natural lip-sync using Wan2.2-S2V in ComfyUI. Covers model setup, audio-to-video pipeline, and workflow configuration for realistic talking head video.
Comments
Sign in with GitHub to join the discussion.