>>106622555
>NEW UPDATE OF A PERSONAL FAVORITE SDXL CHINK ANIME CHECKPOINT I WILL SHILL IT!

One Obsession v17 SDXL Anime Checkpoint Summary

Model Overview:
This is a versatile anime/2D art generation model that's extremely beginner-friendly. It produces high-quality images without requiring LoRAs or complicated prompts.

Version 17 Key Features:
Improved anatomy Better hands and feet rendering without LoRAs
Enhanced colors More vivid and vibrant color palette
Reduced 3D effect Less 3D appearance compared to v14, more anime-style
Beginner-friendly Simple tags produce high-quality results
Red-themed examples V17 examples feature red themes commemorating historical events

Style Flexibility:
The model can produce various styles:
2D flat anime (`flat color, anime screenshot`)
Semi-realistic (`realistic, photorealistic`)
Enhanced lighting (`chiaroscuro, high contrast`)
Dark/moody (`dark, chiaroscuro`)

Recommended Settings:
Steps: 25-35
CFG Scale: 3-6
Sampler: Euler a/Euler
Resolution: 768x1344, 832x1216, 768x1280, 704x1408, 1024x1536

Basic Prompt Structure:
Positive: `masterpiece, best quality, amazing quality, very aesthetic, absurdres, newest, depth of field, highres`
Negative: `worst quality, normal quality, bad anatomy, extra fingers, watermark, low quality`

https://civitai.com/models/1318945?modelVersionId=2177349