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

Short answer: Unless it's a model/feature not available yet on native, you shouldn't.

Long answer: Due to the complexity of ComfyUI core code, and my lack of coding experience, in many cases it's far easier and faster to implement new models and features to a standalone wrapper, so this is a way to test things relatively quickly. I consider this my personal sandbox (which is obviously open for everyone) to play with without having to worry about compability issues etc, but as such this code is always work in progress and prone to have issues. Also not all new models end up being worth the trouble to implement in core Comfy, though I've also made some patcher nodes to allow using them in native workflows, such as the ATI node available in this wrapper. This is also the end goal, idea isn't to compete or even offer alternatives to everything available in native workflows. All that said (this is clearly not a sales pitch) I do appreciate everyone using these nodes to explore new releases and possibilities with WanVideo.

1. Clone this repo into custom_nodes folder.
2. Install dependencies: pip install -r requirements.txt or if you use the portable install, run this in ComfyUI_windows_portable -folder:

python_embeded\python.exe -m pip install -r ComfyUI\custom_nodes\ComfyUI-WanVideoWrapper\requirements.txt

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

fp8 scaled models (personal recommendation):

https://huggingface.co/Kijai/WanVideo_comfy_fp8_scaled

Text encoders to ComfyUI/models/text_encoders

Clip vision to ComfyUI/models/clip_vision

Transformer (main video model) to ComfyUI/models/diffusion_models

Vae to ComfyUI/models/vae

You can also use the native ComfyUI text encoding and clip vision loader with the wrapper instead of the original models:

GGUF models can now be loaded in the main model loader as well.

---

Supported extra models:

SkyReels: https://huggingface.co/collections/Skywork/skyreels-v2-6801b1b93df627d441d0d0d9

WanVideoFun: https://huggingface.co/collections/alibaba-pai/wan21-fun-v11-680f514c89fe7b4df9d44f17

ReCamMaster: https://github.com/KwaiVGI/ReCamMaster

VACE: https://github.com/ali-vilab/VACE

Phantom: https://huggingface.co/bytedance-research/Phantom

ATI: https://huggingface.co/bytedance-research/ATI

Uni3C: https://github.com/alibaba-damo-academy/Uni3C

MiniMaxRemover: https://huggingface.co/zibojia/minimax-remover

MAGREF: https://huggingface.co/MAGREF-Video/MAGREF

FantasyTalking: https://github.com/Fantasy-AMAP/fantasy-talking

MultiTalk: https://github.com/MeiGen-AI/MultiTalk

EchoShot: https://github.com/D2I-ai/EchoShot

ReCamMaster:

WanVideo2_1_recammaster.mp4

TeaCache (with the old temporary WIP naive version, I2V):

Note that with the new version the threshold values should be 10x higher

Range of 0.25-0.30 seems good when using the coefficients, start step can be 0, with more aggressive threshold values it may make sense to start later to avoid any potential step skips early on, that generally ruin the motion.

WanVideo2_1_00004.1.mp4

Context window test:

1025 frames using window size of 81 frames, with 16 overlap. With the 1.3B T2V model this used under 5GB VRAM and took 10 minutes to gen on a 5090:

WanVideo_long.mp4

---

This very first test was 512x512x81

~16GB used with 20/40 blocks offloaded

WanVideo2_1_00002.mp4

Vid2vid example:

with 14B T2V model:

WanVideo2_1_T2V_00062.mp4

with 1.3B T2V model

WanVideo2_1_T2V_00061.mp4