SDXL Unstable Diffusers ☛ YamerMIX

Updated: Oct 04, 2023
Checkpoint Trained
Jun 14, 2023
Base Model
SDXL 1.0
Epochs: 100
Usage Tips
Clip Skip: 1
See Reviews
Yamer's Avatar
Joined Oct 04, 2023

Checkpoint Type: SDXL/Zero Limitations/Everything/Multi Style

Support me on Twitter: @YamerOfficial

Discord: yamer_ai

Welcome to the "SDXL Unstable Diffusers - V8 HEAVEN'S WRATH", the unstable models are a collection of checkpoints where limitations cease to exist and creativity knows no limits. This checkpoint defies restrictions and empowers you to generate anything your imagination desires; 𝐘𝐨𝐮 𝐡𝐚𝐯𝐞 𝐭𝐡𝐞 𝐩𝐨𝐰𝐞𝐫 𝐭𝐨 𝐬𝐡𝐨𝐰 𝐛𝐞𝐚𝐮𝐭𝐢𝐟𝐮𝐥 𝐚𝐫𝐭! 🖌️

As the name suggests, the results obtained from this checkpoint exhibit an inherent unpredictability. You may stumble upon the pinnacle of perfection with a specific prompt or encounter a delightful chaos of visuals. Embrace the adventure and discover for yourself what works best. Good luck in your artistic endeavors fellow creator!

This checkpoint was created with the intent to be unstable, with that I mean that it was created using a huge variety of models, the Ustable Diffusers models are both trained and merged, trained+merged are the models that will usually give you the best results in any type of style because of the huge quantity of styles it can reproduce.

All versions of the model except Version 8 come with the SDXL VAE already baked in, another version of the same model with the VAE baked in will be relased later this month; Where to download the SDXL VAE if you want to bake it in yourself: Click here.

The primary goal of this checkpoint is to be multi use, good with most styles and that can give you, the creator, a good starting point to create your AI generated images and start your own career in the AI world.

^w^ It works best with:

  • LoRA: XL YAMER'S STYLE ♠️ Princeps Omnia LoRA

  • Restore Face: It is not needed to generate high quality images (Not used in the current showcase)

  • Clip skip: 1-2 (The model works well with clip skip set to two)

  • SDXL Refiner: Not needed

  • Checkpoint tested with: A1111

Fields where this model is better than regular SDXL1.0:

  • No embedding needed

  • Better variety of style

  • The refiner is not needed

  • Less AI generated look to the image. (When creating realistic images for example)

  • No face fix needed

  • Better saturation, overall quality, texture and tonality

  • Less blurry images

  • No adetailer required (Just use hires fix)

Improvements over the last version:

  1. Style is stronger. (More defined and unique)

  2. Better with SFW and NSFW. (A nsfw LoRA is recommended for specific nsfw results)

  3. Better coloration of the images.

  4. More variability.

  5. Better close ups.

  6. Less image artifacts.

  7. Sharper and darker lines.

  8. A better look to abstract concepts.

1) Reality is often mischievous, create your own.
2) Sealed and buried for all time is the key to mankind's future. It will take you a million light years from home. But will it bring you back?.
3) Be chaotic, be creative, never stop improving and never put a limit to yourself.


P͓̽e͓̽r͓̽s͓̽o͓̽n͓̽a͓̽l͓̽ ͓̽R͓̽a͓̽t͓̽i͓̽n͓̽g͓̽:͓̽

  • Quality: 9/10

  • Stability (Variety): 6/10, the lower = the more variety with one prompt

  • Flexibility: 8.5/10

  • CFG resistance: 30/30

  • Style strenght: 7.5/10

  • Style Variety: 8.5/10

Recommended settings:

Image Quality: 1024x1024 (Standard for SDXL), 16:9, 4:3

Steps: 35-150 (under 30 steps some artifact may appear and/or weird saturation, for ex: images may look more gritty and less colorful).

  • Hires upscaler: 4xUltraSharp

  • Hires upscale: The only limit is your gpu (I upscale 1.5 times the base image, 576x1024)


To achive a specific nsfw result I recommended to use a SDXL LoRA, this is not a nsfw focused model. (SD 1.5 LoRas don't work with SDXL models)

Video link: "𝕴𝖙 𝖎𝖘 𝖜𝖗𝖎𝖙𝖙𝖊𝖓, 𝖎𝖙 𝖎𝖘 𝖙𝖍𝖊𝖎𝖗 𝖙𝖎𝖒𝖊 𝖙𝖔 𝖌𝖎𝖛𝖊 𝖕𝖊𝖓𝖆𝖓𝖈𝖊 - 𝖎𝖋 𝖞𝖔𝖚 𝖈𝖔𝖓𝖙𝖎𝖓𝖚𝖊 𝖞𝖔𝖚 𝖜𝖎𝖑𝖑 𝖇𝖗𝖎𝖓𝖌 𝖉𝖔𝖜𝖓 𝖙𝖍𝖊 H𝖊𝖆𝖛𝖊𝖓'𝖘 W𝖗𝖆𝖙𝖍."