Create native‑1080p videos from text or images in seconds with Alibaba Cloud’s breakthrough Mixture‑of‑Experts model, Wan 2.2.
From indie storytellers to global brands, everyone’s shooting blockbuster‑quality clips in minutes with Wan 2.2.
Wan 2.2 Workbench
Please sign in to use this feature
Photo
Prompt
Ready to Create Magic ✨
Upload your photos and enter a prompt to generate an amazing AI video. Your creation will appear here once processing is complete.
Watch sample clips and see how simple prompts turn into breathtaking film‑quality footage.
No editing skills required—Wan 2.2 turns your idea into cinema‑quality footage in moments.
Write a scene or drop a storyboard frame
Tune length, aspect and style with smart LoRA controls
Render and download your HD clip
Generate crisp, true‑HD videos ready for publishing and post‑production.
Specialized expert networks boost capacity and visual fidelity without extra compute—making Wan 2.2 the world’s first open‑source MoE video model.
Start from a prompt, a still frame or both; Wan 2.2 supports prompt‑only or frame‑guided generation at 24 fps.
Run locally on a single RTX 4090 or use our cloud API from just $0.02 / sec for 480p and $0.10 / sec for 1080p.
""Wan 2.2 lets me prototype ad spots in hours instead of weeks—game‑changer!" — Beta user"
"I was genuinely moved by the video. Wan 2.2 brings memories to life in such a beautiful way."
"Using Wan 2.2 was an emotional experience. The results felt deeply personal and full of warmth."
From indie storytellers to global brands, everyone’s shooting blockbuster‑quality clips in minutes with Wan 2.2.
For starters and hobbyists that want to try out.
$20.9$14.9/per month
1200 credits per year
All tools available
Email support
Community support
For enthusiasts that want to try out.
$34.9$27.9/per month
2400 credits per year
All tools available
Email support
Community support
For professionals that want to try out.
$53.9$43.9/per month
4800 credits per year
All tools available
Email support
Community support
Secure Payment:
480p, 720p and native 1080p at up to 24 fps.
A 5‑second 1080p clip typically renders in under 40 seconds on an RTX 4090.
Yes. Model weights are open‑sourced on Hugging Face; a single high‑end GPU like the 4090 is recommended.
Yes. Wan 2.2 weights are released under Apache‑2.0, enabling commercial projects.
You can check the pricing on the pricing page.
Yes—you can upload a key frame or storyboard to steer motion and composition.
Input: JPG, PNG, WebP for images. Output: MP4 format with H.264 encoding for maximum compatibility.
Videos can be generated up to 10 seconds long. Longer clips can be created by stitching multiple generations together.
You can try the free trial without signup. For continued use and higher quality generations, create a free account.