Fauxlume XL is based on One Obsession V16 NoobAi. I merged and finetuned it with some LoRAs of my choice.
This model is my approach to create nice 2.5D to Semi-realistic Anime style, but it's not stop there. Adding other styles will also give a distinct and appealing aesthetic.
v1.0:
Initial Release
v2.0:
Merge with Raehoshi illust XL to add more dataset to the model. It can recognize many character from popular anime and game up to October 2025
Slightly reduce color brightness, giving more vivid color
- - - - -
To enhance semi-realistic style, add:
(realistic, photorealistic, realistic shading)
To get a more classic anime look, add:
(anime screenshot), (anime coloring)
To reduce the semi-realistic style, add to negative prompt:
(photorealistic, realistic, 3d, photography)
Recommended Settings
Positive prompt: masterpiece, best quality, very aesthetic, absurdres, sensitive,
Negative prompt: lowres, bad quality, worst quality, bad anatomy, bad hands, sketch, jpeg artifacts, ugly, poorly drawn, censor, signature, watermark,
CFG: 5 ~ 7
Step: 20 ~ 30
Sampler: Euler a / DPM++ 2M Karras
Recommended Resolution
768x1344 (9:16)
832x1216 (2:3)
896x1152 (3:4)
1024x1024 (1:1)
1152x896 (4:3)
1216x832 (3:2)
1344x768 (16:9)
Use ADetailer and Upscale to get the best result.
Hires upscale: 1.5
Hires steps: 20,
Hires upscaler: 4x-AnimeSharp,
Denoising strength: 0.3 ~ 0.5