This all-in-one UI is designed to be an intuitive workflow that works from left to right. Just load your models, enable your features, tweak the settings, and you're good to go. It's a modular studio for quick generations of images and animations, and the best part is that everything is done with switches! No need to rewire or move to a different workflow. Every feature only requires to be switched on with a single click.

-Txt2Img, Img2Img, In/outpaint.

-Txt2Vid, Img2Vid, Vid2Vid.

-PuLID, for face swapping.

-IPAdapter, for style transfer.

-ControlNet.

-Face Detailing.

-Upscaling, both latent and model Upscaling.

-Background Removal.

What's great is that due to the modular design of the workflow you can combine any number of these features for fantastic results. Want Keanu Reeves face as a king sitting on a throne in the style of Van Gogh? Just enable PuLID and IPAdapter with the correct references and you're good to go. Oh. The face didn't come out right? Just enable the face detailer as well and watch the magic happen.

INSTALLATION: Due to the sheer number of features in this workflow there is quite a bit to download but don't worry, I'll link everything here so It's as easy as possible.

LOADERS:

-SDXL Base-  Download sd_xl_base_1.0.safetensors. Place in (ComfyUI\models\checkpoints\)

-LoRA Models- Optional. Place in (ComfyUI\models\loras\)

-SDXL VAE- Download sdxl_vae.safetensors. Place in (ComfyUI\models\vae\)

-ControlNet Models- Download controlnet-union-sdxl-1.0 Place in (ComfyUI\models\controlnet\)

-Upscale Models- Download 4x-UltraSharp.pth. Place in (ComfyUI\models\upscale_models\)

-IPAdapter Models- Download ip-adapter-plus_sdxl_vit-h.safetensors. Place in (ComfyUI\models\ipadapter\) Create folder if you don't have it.

-CLIP Vision Model-  Download CLIP-ViT-H-14-laion2B-s32B-b79K.safetensors. Place in (ComfyUI\models\clip_vision\) I'd Rename it for clarity.

-PuLID Models- Download ip-adapter_pulid_sdxl_fp16.safetensors. Place in (ComfyUI\models\pulid\) Create folder if you don't have it.

IMPORTANT: Download Face-Analysis Models (put in ComfyUI\models\insightface\models\antelopev2\) Download the antelopev2.zip. Unzip the file and place that folder in the \insightface\models\ folder.

-AnimateDiff Models- Download animatediffMotion_sdxlV10Beta.ckpt. Place in (ComfyUI\models\animatediff_models\)

-Face Detailer Models- Should be automatically created when "ComfyUI Impact Subpack" is downloaded from the manager. It should result in a single node called "UltralyticsDetectorProvider".

CUSTOM NODES: There's also quite a bit of nodes that need to be downloaded as well. I'll list them here and link their githubs but you're Comfyui Manager should be able to get most of them by utilizing its "Install missing nodes" feature.

-ComfyUI-Manager

-ComfyUI_IPAdapter_plus

-comfyui_controlnet_aux

-ComfyUI-AnimateDiff-Evolved

-ComfyUI Impact Pack

-ComfyUI Impact Subpack

-rgthree-comfy

-PuLID_ComfyUI

-comfyui-dream-project

-WAS Node Suite (Revised)

-ComfyUI-VideoHelperSuite

-ComfyUI Essentials

That should be it. Please let me know if you run into any troubles with the workflow or the installation. I'd like to remove as many pain points as possible. I've tested everything on an 8 gig vram 3070 and it works although I haven't been able to test the Vid2Vid extensively. Feedback on that would be greatly appreciated. Also, if you run into issues. Use Gemini at AIstudio.google.com. It's an incredible life saver. Just upload the workflow JSON and have it analyze it and help you.