Wan 2.2
Video Model
Image to Video (Plus)
Text to Video (Plus)
Text to Image (Plus)
Text to Image (Flash)

Start with WAN 2.2 – Download Full Package Now
Get Wan 2.2 Model Package
Models | Download URL | Note |
---|---|---|
TI2V 5B | Download Wan 2.2 TI2V 5B Model | Hybrid version supporting both text-to-video and image-to-video, a single model meets two core task requirements |
T2V A14B | Download Wan 2.2 T2V A14B Model | Generates high-quality videos from text descriptions, with 480P and 1080P support |
I2V A14B | Download Wan 2.2 I2V A14B Model | Converts static images into dynamic videos, with 480P and 1080P support |
TI2V 5B Diffusers | Download Wan 2.2 TI2V 5B Diffusers Model | |
T2V A14B Diffusers | Download Wan 2.2 T2V A14B Diffusers Model | |
I2V A14B Diffusers | Download Wan 2.2 I2V A14B Diffusers Model |
Get Wan 2.2 ComfyUI Workflow
Workflows | Download URL | Note |
---|---|---|
Wan2.2 Video Generation ComfyUI Official Native Workflow Example | Download Wan 2.2 ComfyUI Official Workflow | Official usage guide for Alibaba Cloud Tongyi Wanxiang 2.2 video generation model in ComfyUI |
Get Wan 2.2 RunComfy Workflow
Workflows | Download URL | Note |
---|---|---|
Wan 2.2 | Open-Source Video Gen Leader | Download Wan 2.2 RunComfy Workflow | Available now! Better precision + smoother motion. |
Introduction to Wan 2.2 Video & Image Generator
Wan 2.2 marks the next evolution of Wan AI's multimodal generative model series, enhancing creative workflows with greater precision and flexibility. Building upon Wan 2.1, Wan 2.2 introduces refined image and video generation, upgraded special effects, and smarter creative assistance tools. Wan 2.2's powerful generation tool offers creators—from illustrators to game developers—an AI-powered platform to produce stunning images, smooth animations, and immersive effects. With features like real-time parameter control, smart composition suggestions, and cross-modal output capabilities, Wan 2.2 enables rapid, high-quality content generation tailored to diverse professional needs. Wan 2.2 will soon be launched on the RunComfy platform. Its predecessor, Wan 2.1, is already available on the platform and ready to try out immediately.
Features of Wan 2.2 for Creative Generation

Image Generation: More Refined, More Controllable
With Wan 2.2, you gain access to advanced image precision—perfect for projects demanding visual depth and style nuance. Enjoy improved detail for elements like skin, fabric, and landscapes while customizing brushstroke intensity and color saturation. Wan 2.2 also brings you smart composition suggestions to optimize layouts. Whether you are crafting concept art or polished visuals, Wan 2.2 helps you maintain creative control and boost your image generation workflow.

Video Generation: Smoother Motion Effects
Say goodbye to frame flickering and jarring transitions. Wan 2.2 enhances your video content with improved temporal consistency, seamless scene shifts, and support for longer, high-resolution clips. Ideal for short-form creators and storyboard artists, Wan 2.2 helps produce smooth animated sequences quickly. Tap into Wan 2.2 for natural motion effects and visual consistency in video production at any scale.

Special Effects: More Realistic and Diverse
Transform your creative output with realistic lighting, dynamic particles, and stylized filters in just a few clicks. Wan 2.2 enhances special effects by providing global illumination, smarter smoke and fire effects, and automatically matched filter suggestions. Perfect for game developers or multimedia artists, Wan 2.2 streamlines effects generation while maximizing visual impact and production efficiency.

Cross-Modal Creation: Seamless Integration
Wan 2.2 bridges image and video workflows effortlessly. Convert static images into dynamic scenes—like making characters smile or trees sway. Or, extract sharp stills from video for branded visuals. Wan 2.2 ensures style consistency across formats, making it an all-in-one solution for ad teams and visual storytellers aiming for cohesive, cross-platform content with rapid turnaround.

Creative Assistance: Possibly Smarter Tools
Unlock intelligent support with Wan 2.2's advanced creative assistance features. From real-time generation previews to expanded preset templates in anime, realism, and ads, you get tools that flex with your ideas. Wan 2.2 also fine-tunes LoRA models for custom styling support. Whether you're prototyping or finalizing designs, Wan 2.2 delivers fast, flexible creative aid every step of the way.
Prompt Formulas for Better Video Control

Basic Formula: Simple Prompts for Creative Inspiration
Perfect for beginners and those using AI video as a source of imagination. This formula allows free-style prompts that generate highly creative outcomes. The formula is: Prompt = Subject + Scene + Motion. *Subject*: The main object of the video, such as a person, animal, object, plant, or a fictional being. *Scene*: The environment the subject exists in, including background or foreground, real or imagined. *Motion*: Includes both the subject’s movement and any scene dynamics, from stillness to dramatic actions.

Advanced Formula: Rich Details for Enhanced Storytelling
Ideal for experienced users seeking cinematic expression. This formula enriches subject, scene, and motion with fine-grained controls, enabling more vivid, expressive videos. The formula is: Prompt = Subject (Description) + Scene (Description) + Motion (Description) + Aesthetic Control + Stylization. Add details like: 'a young Miao girl in ethnic dress,' or 'a fallen angel from another world with fragmented wings.' Include lighting, perspective, and camera language: 'soft backlight, wide shot, tracking shot.' Stylization terms like 'cyberpunk,' 'line-art illustration,' or 'wasteland style' bring visual identity to life.

Image-to-Video Formula: Describe Motion and Camera
Tailored for workflows where the image has already set the subject, scene, and style. The prompt focuses on describing the motion and camera movement.The formula is: Prompt = Motion + Camera Movement. *Motion*: Describe actions or state changes of objects in the image, like 'a cat slowly turning its head' or 'a child running quickly.' *Camera*: Specify dynamics such as 'camera pans left,' 'zoom in,' or 'static camera' to maintain a fixed view.
Example of Lighting Source









Example of Light Quality










Example of Time of Day






Example of Shot Size







Example of Composition





Example of Camera Framing & Style






Example of Color Tone




Example of Motion Control








Example of Stylization












Wan 2.2 on X: News and Community Highlights
Related Playgrounds








Frequently Asked Questions
What is Wan 2.2 and what is it used for?
Wan 2.2 is an upgraded AI tool designed for generating high-quality videos and special effects. It can be used for creative projects such as concept art, animations, and pre-visualizations, offering more control over styles and better realism than previous versions. It includes dedicated models such as 'wan2.2-t2i-plus' for text-to-image, 'wan2.2-t2v-plus' for text-to-video, and 'wan2.2-i2v-plus' for image-to-video generation.
What are the main features of Wan 2.2 compared to Wan 2.1?
Wan 2.2 builds on Wan 2.1 by enhancing resolution quality, style control, video consistency, and LoRA training efficiency. It also introduces seamless cross-modal creation, allowing users to turn text into images ('wan2.2-t2i-plus'), text into videos ('wan2.2-t2v-plus'), or images into videos ('wan2.2-i2v-plus') more naturally.
How does Wan 2.2 improve LoRA training for custom AI models?
Wan 2.2 enhances the LoRA training process with faster model training, few-shot learning (using just 10–20 images), and intuitive visual tuning options. These make it easier for users to create and maintain personalized art styles.
What content can Wan 2.2 generate?
Wan 2.2 supports high-quality video and image generation. It offers improved temporal consistency for smoother video effects and allows users to generate motion from still images or extract frames from videos while maintaining artistic consistency. The platform supports various generation models such as:
- 'wan2.2-t2i-plus': Text to Image
- 'wan2.2-t2v-plus': Text to Video (supports resolutions: 1920×1080, 1080×1920, 1440×1440, etc.)
- 'wan2.2-i2v-plus': Image to Video (supports resolutions: 1080P, 480P)
Who can benefit the most from using Wan 2.2?
Wan 2.2 is ideal for illustrators, animators, game developers, filmmakers, and advertising professionals. Its refined generation tools and smart composition features help streamline creative workflows and expand multimedia possibilities across industries.
Does Wan 2.2 work on mobile devices?
Yes, Wan 2.2 will be accessible via the Runcomfy website, which is fully compatible with mobile browsers. This makes it convenient for users to create and review content on the go.
What types of output quality can I expect from Wan 2.2?
Users can expect higher-resolution images and videos with enhanced realism in textures, motion effects, and lighting. Wan 2.2 also provides more artistic control, with features like brushstroke adjustments and smart effect matching.
How does Wan 2.2 handle special effects?
Wan 2.2 supports more diverse and realistic special effects, including advanced lighting, particles like smoke and fire, and a growing library of stylized filters. It can also suggest effect combinations automatically based on your creative prompts.
How can I give feedback or report issues with Wan 2.2?
If you experience issues or have suggestions for Wan 2.2, you can send feedback directly to hi@runcomfy.com. User input plays a crucial role in refining the experience and adding helpful new features.
How do I use the text-to-video, text-to-image, or image-to-video models in Wan 2.2?
To use specific generation types, select the 'model' field accordingly:
- Text to Image: use 'wan2.2-t2i-plus'
- Text to Video: use 'wan2.2-t2v-plus'
- Image to Video: use 'wan2.2-i2v-plus'
Make sure to specify supported resolutions in the 'parameters.size' or 'parameters.resolution' fields.
Can I control the resolution of the generated videos in Wan 2.2?
Yes. For 'wan2.2-t2v-plus', supported 'parameters.size' values include:
- 1920×1080, 1080×1920, 1440×1440, 1632×1248, 1248×1632, 480×832, 832×480, 624×624
For 'wan2.2-i2v-plus', valid 'parameters.resolution' values are:
- 1080P, 480P