You can get this model, along with many others, by joining The Tinkerer on Whop. It’s one monthly membership, and you’ll also get early releases, private tools, and a bunch of extra stuff.
CyberIllustrious blends the artistic flair of Illustrious-XL with a big dose of CyberRealistic’s style. You still get all those cinematic, dreamy strokes, but now they come through with crisp, realistic detail.
If you want that painterly, elegant look but still want your images to feel sharp and lifelike, this is your model. It’s great for editorial vibes, soft glam, or even fantasy art that looks like a real photo. Basically, it just nails it.
It’s powerful, easy to use, and honestly? Prompting it is seriously fun.
📘 Get the most out of this model with the Illustrious / NoobAI Prompting: Master Guide.
⚙️ Recommended Settings
Sampling method: DPM++ SDE Karras / DPM++ 2M Karras - SGM Uniform / Euler a
VAE: is already Baked In
Sampling steps: 30+ Steps
Resolution: 896x1152 / 832x1216
CFG: 4-5About (DMD2)
The special DMD2 edition of CyberIllustrious is created for faster, cleaner generations while requiring fewer steps.
DMD2 takes the balanced improvements of the standard release and optimizes them for speed, precision, and responsiveness, making it ideal for both SFW and NSFW workflows.
Expect tighter results, smoother control, and a model that reacts beautifully to the classic Illustrious prompting style.
Recommended Settings
Sampler: LCM
Schedule Type: Exponential
Sampling Steps: 8 (12 for extra fine detail)
CFG Scale: 1.5
⚡ These settings are optimized for DMD2 and are different from my standard Illustrious workflow!
🧠 Prompting Made Easy
Positive Prompt:
masterpiece, best quality, high quality, ultra-detailed, realistic, ... rest of promptNegative Prompt examples (optional):
cartoon, illustration, anime, painting, CGI, 3D render, low quality, watermark, logo, labeldeformed, deformed face, low quality, bad quality, worst quality, (drawn, furry, illustration, cartoon, anime, comic:1.5), 3d, cgi, extra fingers, deformed teeth, watermark, logo🩺 Optional ADetailer Settings
Adetailer model: face_yolov9c.pt
If you only want the main face being refined set 'Mask only the top k largest' to 1.🔗 Resources
Backup location: HuggingFace
☕ Support the Madness
Liking what you see? Grab me a digital espresso (or a couple!). It keeps the creative energy going and helps bring more cool prompts and models your way.
⚠️ Heads Up
This model can spit out some mature stuff. Use it wisely, and please, no cursed creations—we’ve really seen enough of those.
💬 Want Better Prompts?
Use this custom ChatGPT to auto-generate polished prompts for CyberIllustrious:
🔗 [Try it now on ChatGPT]
Description
Further refined the model, fixed some mistakes (and probably introduced new ones in the process).
FAQ
Comments (19)
any tips or general theory about lora compatability? i noticed that illustrious loras struggle with this checkpoint as opposed to another one i found, which somehow does kind of well with realism even though it is illustratey on the tin. https://civitai.com/models/1161909/knk-noobicomp The creator of that checkpoint says in the description something about the way they built it for greater Lora compatability. any thoughts on the science behind this? and if there's a way to give your illustrious checkpoint more "room" for lora compatability, do you think it's something you'd want?
include gelbooru datasets in the future?
Really solid- you definitely pulled off a realistic Illustrious model.
Although I'd recommend dpmpp_2m + sgm_uniform
Excellent recommendation
I came here to say the exact same thing.
@panorios I use DPM++ 2M Karras mostly all the time. SGM Uniform is very good alternative for Karras
As a rule, Illustrious models do not go with clip skip 2. Is there a special reason why you recommend to use it?
When Sdxl dropped 1,5 years ago, clip skip has nothing to do with generation. 100% same on 1 and 2.
I copied/paste te settings and forgot to remove the steps section. I fixed it now.
@Cyberdelia I dont know Comfy, but in Forge and i think Auto1111 clip1 and 2 produce exatly the same results.
@aidynnkermit497 I know. i also use Forge and there is no difference between 1 or 2.
@Cyberdelia in Comfy it produces nothing but noise when setting "clip skip" to 1, while a value of 2 works perfectly fine
I was a bit harsh earlier, its definitely good in this state already. Lighing and shadows are miles better, than Pony, i didnt even prompt them.
I removed the score tags, the rest prompts remained the same, as in Pony, the results are good, but need refining of course.
It struggles with simple sex position, like sidewiew doggystyle. It gives POV view, but at least there are penetration always. Side view seems tricky, because in Pony it gives sometimes full solo girls, or body horror. My prompt is precise, booru tags.
I use a lot detailer, realistic prompts, realism detailers, with these, its almost as realistic as Pony.
Your samples dont give justice to the model, it can be way more realistic, thatswhy i didnt try earlier.
Allright, keep up the good work.
And if somebody has tips for proper sex poses, share them, thank you.
You’re absolutely right about the example images - they were thrown together faster than a microwave meal! I’ll definitely fix them up, because I can see how they might scare people away from giving this a try.
And about your question on ‘sex poses’ - you’d be surprised how many books exist that can teach you more positions than the average yoga class. So if that’s what you’re after, there’s a whole library just waiting to enlighten you!
that part is so true: the samples don't give justice to the actual image quality of the checkpoint. I didn't test it before for the same reasons. I'll try to write a review later on going more indepth but its a very impressive work for being at this very early stage.
@ElectricDreams I’m currently feeling like a deflated balloon because some harsh words were thrown at my brilliant example images - guess they weren’t ready for this level of artistic greatness! ;)
@Cyberdelia what I meant is that samples look cartoony alla pony style whilst the checkpoint can do photorealistic (something that pony can't really do on my book). Or al least has the potential to be great at photorealism. But maybe that isn't the goal of the checkpoint. That was pretty much it. Your images are awesome in a cartoony way.
@ElectricDreams I know, I was just making a joke. You’re right about those images :)
@Cyberdelia i know i know.