NokiMo
Innovate Futures @ Benji
Innovate Futures @ Benji

patreon


Wan2.1 With ComfyUI Text2Video And Image2Video Locally

Video : https://youtu.be/bLnr1T7DyTI

WAN 2.1 AI: Revolutionize Video Creation with Alibaba's Open-Source Model!

Discover WAN 2.1, Alibaba's latest open-source AI model for text-to-video and image-to-video generation! In this tutorial, learn how to set up and run WAN 2.1 locally in ComfyUI, exploring its advanced features like dynamic camera angles, video editing, and multi-image referencing. Perfect for creating cinematic car animations, futuristic jet fighters, and more—all with enhanced motion and detail!

Who is this content suitable for?

AI developers, video creators, and tech enthusiasts looking to leverage cutting-edge AI models for professional-grade video generation. Ideal for ComfyUI users with mid-to-high-end GPUs or those exploring open-source AI tools for creative projects.

Why does it matter?

WAN 2.1 pushes the boundaries of AI video generation, offering smoother motions, richer details, and dynamic camera movements. Whether crafting ads, cinematic clips, or experimental animations, this model empowers creators to produce high-quality content without costly subscriptions.

https://x.com/Alibaba_Wan

https://wanxai.com/

https://tongyi.aliyun.com/wanxiang/

https://github.com/Wan-Video

https://huggingface.co/Wan-AI

https://github.com/kijai/ComfyUI-WanVideoWrapper

https://huggingface.co/Kijai/WanVideo_comfy/tree/main

Workflow in tuturoial attached below. Have fun :)

Wan2.1 With ComfyUI Text2Video And Image2Video Locally

Comments

Hey man , hows doing :) Yes, it can adjust the width and height to portrait, then it good. For ratio, i have try in portrait , 3:4, 3:6, 9:16. And it work too.

Benjamin Law

hi. thank you for great content. I understand that the model is optimized for horizontal videos are there a workarounds such as adjusting the width/height settings in ComfyUI or rotating input images. Would you recommend any specific method to achieve 9:16 (portrait) aspect ratio videos with WanVideoModel? Is there a way to modify the workflow or model settings to better support vertical video generation?

modiin wrestler

You can swlect Sadpa if you don't have Sege att in Windows. sage attention need additional libraries install, i will find a link of guide send you later. I remember I saw I web talk about this.

Benjamin Law

I got "Can't import SageAttention: No module named 'sageattention'" straightly using kijai , any thought and pointer how I resolve it? Did I miss something?

Vito Kwan


Related Creators