Illustrious XL-Modular WF v1.0

Illustrious XL-Modular WF v1.0

Illustrious XL, developed by Onoma AI, has rapidly emerged as a prominent AI model in the realm of text-to-image generation, particularly within the anime and illustration communities. Built upon the Stable Diffusion XL (SDXL) architecture, Illustrious XL distinguishes itself through its innovative use of Danbooru-style tagging and a commitment to high-resolution image synthesis.

Illustrious XL allows you to generate anime and cartoon comics, in different styles, using many available LoRAs.  The open-source model was fine-tuned in many new checkpoints focused on 3d illustrations, specific anime, or realism.

In this workflow, you will find three different checkpoints already installed: Illustrious XL v2.0 (the latest original model available), one checkpoint focused on 3D toon illustrations (ilustmixv6), and one aimed at realism (ilxlRealismKlaabu).

These models can also be used for NSFW image generation, and sometimes may output "spicy" images even if you did not want to generate that kind of output, so be careful!

The workflow will let you generate txt2img and img2img outputs, it has the following modules:  HiRes Fix, Ultimate SD Upscaler, FaceDetailer, and a post-production node.

Also, the generation will stop once the basic image is created (Image Filter node) to allow you to choose whether to continue the workflow with that image or cancel it. This is extremely useful when you generate a large batch of images! If you generate only one image (batch size =1), the workflow will not stop. (In the settings, you can modify the way this node works.)

Here is the first part of the workflow:

(I had to replace the Seed(rgthree) node due to conflicts with the latest ComfyUI update, you can try to use it again as once you update the rgthree-nodes it works without trouble).