Welcome to the planet. I’m SupremoKaioSama — named after Kaiō-sama: humor, rigor, and enough power to make training hurt. I make LoRAs for 3D and anime on Illustrious, Illustrious 2.0, Pony, SDXL, Animagine, and NoobAI. I train alone, so releases are slower on purpose. One model that holds likeness and prompt control beats ten that only look good in the preview grid. Expect character LoRAs first (Dragon Ball and classic/retro anime), plus the odd style or experiment. Pages include triggers, recommended strength, and honest notes when something is still wonky. SFW and +18 both appear here; use filters if you only want one lane. Want to help? Generate with the models, post your images, leave a short comment, and hit like. Buzz tips fund the next dataset. If a LoRA misfires, read the version notes first. Clip skip, weight, and base checkpoint matter more than people think. Better recipes get added when I find them. More 3D and anime characters and cleaner 3D-friendly versions are in the queue. Train hard, prompt smarter. Power is not in the model. It’s in how you use it.