This little baby is an all-in-one, waifu-powered workflow!
✨Smooth Wan Shooter has been realeased, featuring SmoothMix! Get the model's here: https://civarchive.com/collections/7237154
Thank you @DigitalPastel for your help! Give them a follow and check out their work!
The original version 1.0 and 1.2 both use my paired-down Wai-Illustrious v.15 1911A1, with special tuning for image to video, to create an input image for Wan 2.2. Everything has been adjusted to highlight the anime and hentai style! 💖
To Use: Download the .zip, unpack the JSON, and drop it into ComfyUI!
Please only use the downloaded Workflow version, as some of the videos are from my development and testing, and I would hate for you to have to recalibrate it! The videos are great examples of successful prompts, and there's even a few errors if you look closely in ComfyUI! 😰
Detailed Readme, Markdown Labels, and Prompt Tips included in the JSON!
It creates detailed and anatomically correct 1024x1024 images to convert into 480p or 720p videos!
The Wai-Fu Wan Shooter edition uses the Wan 2.2 14B I2V GGUF checkpoints with lightning LoRAs.
Here is the NSFW UMT Text encoder on Huggingface (I'm on there too) from NSFW API: https://huggingface.co/NSFW-API/NSFW-Wan-UMT5-XXL/tree/main
The Hiromitsu Hentai Cannon uses @DigitalPastel 's SmoothMix NSFW WAN I2V checkpoints: https://civarchive.com/models/1995784?modelVersionId=2260110
I used Takeda Hiromitsu Styling as a base for my image testing, using the excellent LoRA by @Goofy_Ai https://civarchive.com/models/12542/takeda-hiromitsu-style-illustriouspony15-or-goofy-ai
For video testing, I used @DigitalPastel 's SmoothMix XXX animations: https://civarchive.com/models/12542/takeda-hiromitsu-style-illustriouspony15-or-goofy-ai
As well as @ai_build_art 's Bouncy Walk: https://civarchive.com/models/1361346/bouncy-walk-wan-i2v-14b?modelVersionId=2261902
⚠️ DISCLAIMER ⚠️
I take no accountability or responsibility for however you choose to use this workflow and AI tool! I am officially telling you, Do NOT use this tool for anything illegal or disgusting, and I think you can imagine what a fellow degenerate like me would consider actually disgusting... 🤢. It is completely uncensored, so this is your warning!
Description
First release All-In-One workflow including detailed Readme and Markdown Notes
FAQ
Comments (13)
soooooooooo sexyyyyyyyyy
Thank you, sweetie!
@pixelladies45 you're very welcome
Man ... the Wan 2.2 14B I2V GGUF checkpoints is really not the ideal for anime generation. It's incredible how your results look
Oh! Thank you, cutie! 💖
Haha yes, you're right about Wan 2.2! 😆 When I was making this workflow, Jessica Rabbit kept morphing into a realistic woman! Don't get me wrong, that's pretty hot and all, but NOT what I was trying to do at the time 🤭
I had to adjust my LoRAs and prompts to steer towards more anime styling after that. The "SmoothMixAnime" tag went a long way to help boost the anime levels, thanks to the Smooth Mix Animation LoRAs and GGUF Checkpoints I included as well!
I've also been known to get myself into trouble by making things do stuff they're not supposed to do. Especially on public commercial platforms. The brat in me really comes out when something tells me I'm not allowed to do something. Even more if it's based on censorship. Like okay, bet. 😂
Hi!
I just wanted to say that your WAN workflow is absolutely amazing — it works like a charm 🔥
I’ve had a lot of fun using it and experimenting thanks to your setup.
I had a quick question as well: do you think it would be possible to adapt this workflow to image-to-image (using an input image as a starting point)?
Thanks again for sharing such great work, really appreciate it 🙏
Hi there! 💖 Thanks for the feedback! I'm glad to hear that it's working well, and I've been experimenting with it too 😆
Do you mean image to video? Meaning that it would take an input image and make a video out of it? If so, I'm working on polishing up a release for that right now! I wrote most of the prompting guide today, and several anonymous community members are helping me customize and test it even more! 😸 I hope to release a strict image to video, as well as a first frame last frame version focusing on Wan 2.2 and Smooth Mix GGUF (what I used in the Wai-Fu Wan Shooter and Hiromitsu Hentai Cannon)!
In case you did mean image to image (as in image transformation, editing or inpainting), I may try some Illustrious image to image, Z-image, or Qwen Edit workflows as well! I have less experience with those formats, so I wanted to share what I was best at first!
Also, oh my gosh, I'm so glad for the support! I plan to keep making all kinds of fun SFW and NSFW AI tools! It's been so nice to find this community! 😊
@pixelladies45Hi! 👋
Just a quick heads-up: after updating my ComfyUI setup, I’m running into an issue with this workflow that I didn’t have before.
When using Wan 2.2 GGUF with the standard LoRA loader, I now consistently get this error:
'dict' object has no attribute 'model'
The same JSON was working fine for me a couple of weeks ago, so it seems related to recent ComfyUI / GGUF changes. Totally understand if this setup isn’t supported anymore — just wanted to share in case it’s useful
Thanks again for sharing your work! 🙏
@Blazingtaste13 Thank you so much for telling me!
You may need a Unet GGUF loader node (guide here: https://www.runcomfy.com/comfyui-nodes/ComfyUI-GGUF/UnetLoaderGGUF )! You can probably search the ComfyUI Manager or double click your workflow background and search for "Unet GGUF" in the node search box! 🔍
The new Smooth Wan Shooter comes pre-loaded with both the Unet and the standard "Load Diffusion Model" nodes! ✨
If this doesn't fix it, please tell me! I don't want broken workflows out in the wild! 😉
@pixelladies45 Just a quick update: my issue was caused by a broken ComfyUI venv (NumPy corruption), not the workflow.
A full uninstall + clean reinstall fixed everything.
The final JSON you published is fantastic !! — very clean and works perfectly now. Thanks for sharing!🙏✨
@Blazingtaste13 Great! You're welcome, and I plan to keep sharing! 💖
I'm completely new to ComfyUI and WAN, but this workflow is very well documented and works amazingly well!
Yay, I'm so glad! 💖