Updated: Oct 1, 2026
styleLoaded
Download
1 variant available

21
93
Generation, training and LoRA distribution on Civitai are covered by Civitai’s own license agreement with MiniMax. If you download these weights and run them yourself, your use is instead governed by the MiniMax H3 Community License Agreement, whose grant excludes the European Union, the United Kingdom, the Republic of Korea and the United States of America.
MiniMax H3
This is an experimental follow-up to my Anime IRL LoRA. It aims to render subjects with a fully illustrated style within photorealistic environments. By default, it uses a grayscale palette, but you can also specify colors for individual elements, allowing for selective coloring or even full-body coloring as well. This LoRA is more prompt-sensitive than Anime IRL, so you may need to iterate a bit to get what you're looking for.
Trigger: SKETCHIRL, hand-drawn - you can also add specific art style tags to influence the appearance more toward anime or digital illustration or whatever else you want.
Strength: Start at 1.0, increase if too realistic
Important: Even though the model is biased toward grayscale, if you are aiming for a clean black-and-white look, I would still recommend labeling hair/clothes/accessories as 'white'. Otherwise the colors can drift.
Example Prompt:
SKETCHIRL, hand-drawn, a young woman with long white hair in a white sweater and white pants sits at a kitchen table, slowly stirring a mug of coffee. Morning light comes through the window behind her. She looks down at the mug, then glances toward the window. Static camera, medium shot, natural lightingNotes/Tips:
Avoid using texture or fabric words like 'knit sweater' or 'denim jeans', which cause a more realistic appearance
Sometimes it gives choppy motion. Adding motion/speed tags can help, otherwise just reroll
For multiple subjects, tag each one's style separately

