I created this workflow to serve as both a text-to-image and image-to-image laboratory. It includes several nodes that simplify switching between operations, as well as a Lora stacker that allows for streamlined and condensed additions to the flow. I’m sharing this primarily to keep a record of it here on Civtai, just in case someone finds it useful.
I've included a notes panel that explains how to get started. All the extra nodes can be downloaded through the ComfyUI Manager, and I believe the upscaler can also be accessed directly from the ComfyUI Manager interface.
If you're just trying to make images and feel confused by the many extra options, I've attempted to make the headers and node names more intuitive for those who may not have a programming background. My goal is to help you start creating art without feeling overwhelmed or needing to pay high fees.
This workflow runs smoothly on my laptop; however, your experience may vary. It allows for images to be generated using SD1.5 models, SDXL models, or PONY-based models. I’m currently using an extended version that operates with Flux, though that setup is a bit more complex at the moment. I’ll update this workflow with that version once it feels more streamlined and easier for outsiders to understand.
My SD1.5 model, HICCUP, can be replaced with any SD1.5 model, but I recommend downloading HICCUP from here because it’s worth it.