Wan, visual generation model, Alibaba Group, Hangzhou, Zhejiang, China

[Administrator]

Developer - Alibaba Group

wan.video
wanxai.com

tongyi.aliyun.com/wanxiang

wan-ai.org

wan21ai.com

youtube.com/@Alibaba_WanX

x.com/Alibaba_Wan

huggingface.co/Wan-AI

github.com/Wan-Video

 

[Administrator]

Meet WanX - our next-gen AI model redefining visual creativity!

Feb 20, 2025

 

[Administrator]

Article "Alibaba offers free access to its AI model that can generate realistic video and images"
The company is open sourcing its latest generative AI models.

by Mariella Moon
February 26, 2025

 

[Administrator]

"VACE, All-in-One Video Creation and Editin"

by Zeyinzi Jiang, Zhen Han, Chaojie Mao, Jingfeng Zhang, Yulin Pan, Yu Liu

 

[Administrator]

Wan 2.1 VACE 14B Released - Better And Higher Quality of AI VIdeo Editing Model!

May 19, 2025

Wan 2.1 VACE 14B Released - Better Quality For AI Video Editing - Tutorial Guide

In this video, we dive into the powerful WAN 2.1 V Ace AI video editing model, now available in 1.3B and 14B parameter sizes on Hugging Face, ModelScope, and official API endpoints. Discover its advanced features like object swapping, video inpainting/outpainting, frame interpolation, and masking—all tested locally in ComfyUI. We compare the performance of the Wan 2.1 VACE 14B model with earlier versions, showcasing its superior quality in tasks like character transformation, dynamic scene generation, and resolution upscaling. Whether you're an AI developer, video editor, or tech enthusiast, this video demonstrates how V Ace 14B revolutionizes AI-assisted video production with faster rendering and cinematic detail.

Who is this for?
-AI researchers and developers working with WAN 2.1
-Video editors exploring AI-powered post-production
-Content creators using ComfyUI for dynamic video workflows
-Tech enthusiasts interested in next-gen video synthesis

Why it matters:
The 14B V Ace model marks a leap in AI video editing, enabling high-fidelity transformations, smoother interpolations, and efficient masking—previously limited by smaller models. Its integration with text-to-video workflows and compatibility with tools like CausVid and MovieGen makes it indispensable for professional-grade results.