Move over, Sora—Alibaba has just launched a powerful new open-source video AI called Wan2.1-VACE, and it’s set to change how we create and edit videos.
This new tool is part of Alibaba’s Wan2.1 model family and offers a one-stop solution for everything from video generation to fine-tuned editing—all powered by AI.
What It Can Do
Wan2.1-VACE makes video creation more flexible and easier than ever:
Create from Anything: Turn still photos into animated videos, generate clips from just a text prompt, or even bring characters to life using reference images.
Smart Editing: Edit only the parts of a video you want—change colors, swap poses, or adjust movement without messing up the rest of the scene.
Wider Formats: Want to turn a vertical image into a widescreen video? This model can expand the frame and fill in the blanks with AI-generated visuals.
Advanced Tools: Video Condition Unit (VCU) helps process different types of input (text, photos, videos, and masks) for smooth editing. Context Adapter improves how the AI understands timing and space, leading to more natural video transitions and effects.
Free and Open to All
Unlike other expensive tools, Alibaba has made Wan2.1-VACE free and open-source. Two versions are available:
A 14-billion parameter model for high-performance systems. A lighter 1.3-billion parameter model for everyday use.
You can download both on Hugging Face, GitHub, and ModelScope. This makes it easier for small businesses, educators, and indie creators to produce high-quality content without breaking the bank.
Alibaba’s release shows its strong support for open-source innovation and could push the whole video industry forward—making pro-level tools available to everyone, not just big companies.
This blog post is for informational purposes only