This is a slightly pared down version of my existing T2I workflow, with the addition of being able to switch to Krea 2 Raw or Anima or even Moss with a single toggle switch. You know that you can switch by going from tab to tab with only one click too, but what if your tabs are far afield? Suddenly you're looking for it, it's taking more clicks, besides, you want as few tabs open as possible, what with all the workflows you've got. Just one workflow for FOUR models? That's a win. Well here it is. Want to compare outputs using identical or comparable resolution settings? Just one radio button away.
I also want to say thanks to @catlover1937 for making such good-looking models and putting them up. You have probably seen those Moody checkpoints -- yeah, that catlover1937, now you remember. I used his workflow as a guide to make the additions to my workflow because he published a MINIMAL workflow -- that's all I needed.

KreaAnimaMoss V1.0

Added Moss. Fixed a typo. Used catlover's model guide for providing links to where to download stuff, it will help you. Just open up the workflow and click on a link if there's a model you don't have, and it will download. Also, using the LC Note you can translate the text into any other language.

About MGH T2I Anima / Krea 2 Beta Version

Because it’s built on the previous workflow, most of the familiar pieces are still here:

- First-pass generation with the usual controls

- Upscaler stage

- Detailer array (SEGS-based) plus the LC detailers if you want them

- Bypass logic so you can turn sections on or off cleanly

- Metadata embedding on save

- The same general layout and color coding you’re used to, if you use my workflows

The main new piece is the switching layer. Each mode has its own settings panel (steps, CFG, sampler, size adjust, etc.) and its own LoRA loader. Only the active mode’s loaders and settings are live; the others stay muted. You can flip modes mid-session, adjust the relevant panel, and queue again without touching the rest of the graph.

Also there’s a guided tour built into the workflow (Vision Assistant) that walks through the major groups, and there are notes on the canvas explaining the less obvious parts. If you’ve used the earlier version, the tour is mostly a reminder of where things live plus a quick look at the new mode controls. If you’re new to it, the tour plus the group labels should be enough to get oriented without a full manual.

Recommended starting points are the same as before: keep the detailers off until the base image looks good, freeze the seed, then turn on upscale or detailers as needed. Anima and Krea respond a bit differently to the same prompt, so it’s worth testing a prompt in Turbo first before committing to a longer Raw run, but the thing about Raw is, it's got a lot more freedom than Turbo.

I have had the pleasure of generating with this thing for several hours and put it through the paces of using the detailers for various numbers of faces, using the LC Detailers, and really thinking about the settings and how things are supposed to be. So, while I still call this workflow a Beta, and encourage you to come back and look for updates, and let me know if you run into any errors, for the most part, I think it's in good order. I am trying not to jinx this release but oh well.

Custom Nodes Used:

rgthree-comfy

ComfyUI-KJNModes

ComfyUI Impact Pack

ComfyUI-Easy-Use

ComfyUI-Crystools

efficiency-nodes-comfyui

Comfyroll Studio

ComfyUI_essentials

ComfyUI Impact Subpack

ComyUI Image Saver

ComfyMath

ComfyUI-LogicUtils

JPS Custom Nodes for ComfyUI

Fearnworks Nodes

ComfyUI LC Vision Nodes

ComfyUI LC MaskMaker Nodes

ComfyUI LC123 Nodes