Sign In

CyberRealistic XL

4.6k
97.9k
678.3k
2k
Verified:
SafeTensor
Type
Checkpoint Trained
Stats
2,064
0
Reviews
Published
May 9, 2025
Base Model
SDXL 1.0
Hash
AutoV2
49E4F2939A

Civitai onsite generation enabled

This model was built with two key goals in mind: simplicity and versatility. No need for complicated prompts (unless you're into that sort of thing) - just let your imagination run wild.

Have fun and enjoy the ride!

Personal settings

Sampling method: DPM++ 2M SDE Karras 
VAE: is already Baked In
Sampling steps: 30+ Steps
Resolution: 832x1216 / 896x1152
CFG: 3-5
Clip Skip: 1
Upscale: 1.5 upscale (0.3 - 0.35 denoise)
Upscalers: 4x_NickelbackFS_72000_G / 4x_foolhardy_Remacri / 4x_NMKD-Siax_200k

Example negative prompt:

cartoon, illustration, anime, painting, CGI, 3D render, low quality, watermark, logo, label

or use CyberRealistic Negative


Adetailer settings (if you want to use it, it's not always needed.)

Adetailer model: face_yolo8n.pt
If you only want the main face being refined set 'Mask only the top k largest' to 1.
Inpaint denoising strength: 0.5 (or just let this 0.4)
Use separate width/height: 1024
Use separate steps: 50 -70

Want to buy me coffee? (Buy a cup)