Pretty simple workflow for Krea 2 turbo, similar as my other workflows for Z-image, Flux 2 Klein and Qwen 1.
IMPORTANT!
DemonAlone-JS_Addon-ComfyUI is stuck on version 1.5.3, where I haven't yet added my toggles, and the actual version is v1.9.8. You need to install the nightly version via the ComfyUI manager if your security settings allow it (or change them), or you can go directly into the extension folder and run
git pull origin main from the terminal.
I am tired of beating the dead horse regarding Legacy mode and have decided to try removing Nodes 2.0. However, I didn't anticipate how deep that issue runs. Anyway, here is a version for you.
In theory, everything (except comparison previews) should work in Legacy mode too. But there is a kind of issue: my status toggles will switch nodes into Full mode (too many slots), if you switch the interface from Legacy to Nodes 2.0 with an open workflow. You just need to toggle it via the context menu between Full mode and Compact mode.
Description
Added the Inpaint mode. Some visual and internal changes.
Some internal and visual changes.
FAQ
Comments (2)
I’m not very familiar with ComfyUI, but looking at your workflow, I wondered why you only set up a single "ADetailer" group. Usually, I need to fix several details on the base image such as face(s), eye(s), teeth, hand(s), breast(s), nipple(s), genitals, etc. So, is it complicated to add extra "ADetailer" groups to your workflow to fix multiple stuff in a single run?
Also your LoRA loader doesn't find any of my Krea 2 LoRAs installed using Stability Matrix Model Browser and manually for a couple. Is it safe for me to replace your LoRA loader by one that i use and doesn't suffer from that issue? I m thinking about the LoRA loader from "nd-super-nodes".
@antarek , I added only one adetailer because the detection and sampling works not perfectly. I have a workflow for sdxl-like where a few adetailers are expanded in the workflow, but they use it pretty much with all of them rare exceptions. With a new model like Flux or newer ones, they're usually not as bad and often don't require as much manual inpainting; sometimes they're even more useful to do through manual Inpaint. Adetailer generally just does inpaining via detection system, but you can add as many as you want by making copies of the adetailer subgraph or creating a new one, and also adding additional previews after each one.
Regarding lora loader: yes, you can replace them and trigger words node on any other, like Power Lora. If you run into issues with those modifications, let me know—I'll help troubleshoot those for you.