Create AI Videos with WAN AI
Open Source Leader
Experience WAN 2.2 with revolutionary MoE architecture and 27B parameters. Generate 720p@24fps videos on consumer GPUs with SOTA performance in both text-to-video and image-to-video generation.
Describe your vision - WAN 2.2 will create cinematic videos with SOTA quality
Trusted by Professionals and Creators from leading brands and companies
Join thousands of creators and professionals who trust our AI video generation platform
Endless Creative Possibilities
Explore videos created by our community across different styles and categories
Advanced WAN AI Features for Developers & Creators
Everything you need for professional video creation with WAN 2.2's open-source MoE architecture and consumer GPU compatibility
Introducing WAN 2.2 MoE
WAN 2.2 debuts the first open-source Mixture-of-Experts architecture for video diffusion models. With 27B parameters activating only 14B per step, achieve enhanced efficiency without added computational cost.
SOTA Performance & Consumer GPU Support
WAN 2.2 achieves state-of-the-art performance ranking Top 3 on T2V and I2V leaderboards. Generate 720p@24fps videos on consumer GPUs like RTX 4090 with only 8.19 GB BRAM requirement.
Multi-Language Text Generation
WAN 2.1 is the first video model capable of generating both Chinese and English text. Features advanced Wan-VAE for encoding 1080P videos of any length while preserving temporal information.
Ready to experience open-source SOTA video generation?
Get Started with WANHow to generate videos with WAN AI
Creating professional videos with WAN 2.2's open-source MoE architecture has never been more accessible. Experience SOTA quality on consumer hardware.
Choose Your Generation Mode
Select from text-to-video (T2V), image-to-video (I2V), or hybrid text-image-to-video (TI2V) modes. WAN 2.2 supports all generation types with its unified MoE architecture and advanced VAE system.
Configure & Generate
Set your WAN 2.2 preferences for resolution (480P-720P) and duration. The dual-expert MoE system uses high-noise for structure and low-noise for details, delivering movie-quality visuals with rich textures.
Download & Integrate
Your WAN 2.2 generated video is ready! Download in high quality or integrate using the open-source codebase. Access through ComfyUI, Diffusers, Hugging Face, and ModelScope for seamless workflows.
Ready to create your first WAN 2.2 video?
Join the open-source community using WAN's SOTA video generation technology with consumer GPU accessibility.
Start Creating with WANWhat Creators Say
Trusted by filmmakers, creators, and marketing teams worldwide
- 5.0
Crevas completely transformed my filmmaking workflow. I turned a rough script into a full shot list and generated cinematic-quality video in just one day. What used to take a production team a week, I can now achieve on my own.
Jack Smith, Independent Filmmaker
- 4.9
As a creative director, my biggest challenge was producing consistent style videos at scale. With Crevas’ multi-model integration and one-click export, we produced an entire season of campaign videos 5x faster than our usual workflow.
David Johnson, Creative Director
- 5.0
Our studio delivered its first AI-driven film project using Crevas. The built-in collaboration and version control made remote teamwork seamless, and our client was impressed by both the speed and the cinematic quality of the output.
Kevin Brown, Film Producer
Frequently Asked Questions
Everything you need to know about WAN 2.2's open-source MoE architecture and SOTA video generation capabilities