Updated the Prompting Guide
For business inquires, commercial licensing, custom models, and consultation contact me under [email protected]
Join Juggernaut now on X/Twitter
Juggernaut Ragnarok on RunDiffusion
Juggernaut XI & XII on RunDiffusion
Prompting Guide for Juggernaut Ragnarok by Adam
Prompting Guide by Adam for XI & XII
A big thanks goes to RunDiffusion and Adam, who diligently helped me make it work :) (Leave some love for them ;) )
Hey everyone,
It’s been 8 months since the last version was released here on CivitAI.
Of course, I haven’t been idle during that time . I completed several projects to ensure I’d have the financial means to keep exploring new architectures and possibly do full finetunes on them in the future.
Juggernaut Flux (and its many sub-variants) was a ton of work, but ultimately, I’ve wrapped that chapter up. The training process gave me way too many headaches. To keep my sanity, I spent my spare time working on Juggernaut SDXL with the hope of maybe releasing one final version for you all.
And that day has finally come. :)
Juggernaut Ragnarok has improved in many areas: photorealism, digital painting, poses, hands, feet, and much more.
That said, it’s still an SDXL model, and I don’t recommend comparing it to models like Flux, Reve, or Sora. For example, it still has limitations when it comes to text rendering or faces at a distance.
I recommend using it as part of a pipeline for your projects. Example setup:
FluxDev / Pixelwave / Jug Flux Pro → Juggernaut Ragnarok
A quick personal note about Juggernaut:
Honestly, I don’t know what comes next.
After the release of Sora and similar tools, the open-source image generation space feels a bit dull in comparison.
Nothing has really excited me enough to dive back into training (yes, I’m talking about HiDream too).
So I’m seeing Juggernaut Ragnarok as a kind of farewell, especially since it’s unclear where things are headed with CivitAI in general.
(You can download all Juggernaut versions from HuggingFace, by the way.)
Last but not least:
Have fun with the model, share your creations, and good luck with your projects!
And in case you’re wondering: Yes, you can do anything you want with Juggernaut : merge it, train it, sell the image outputs, etc.
Just a simple shoutout is all I ask. :)
And now, here are the recommended settings:
Recommended Settings(VAE is baked in):
Res: 832*1216 (For Portrait, but any SDXL Res will work fine)
Sampler: DPM++ 2M SDE
Steps: 30-40
CFG: 3-6 (less is a bit more realistic)
Negative: Start with no negative, and add afterwards the Stuff you don´t wanna see in that image.
VAE is already Baked In
HiRes: 4xNMKD-Siax_200k with 15 Steps and 0.3 Denoise + 1.5 Upscale
And now, have fun trying it out. As always, I'm eagerly waiting for your pictures in the Gallery :)
If you liked the model, please leave a Like. In the end, that's what helps me the most as a creator on CivitAI. :)
Last but not least, I'd like to thank a few people without whom Juggernaut XL probably wouldn't have come to fruition:
Dreamlook.AI (Trained 3 Side Sets)
Description
New Side Set with the help of SilasAI6609 (42k Steps)
And another 50k Steps for the Base Model
FAQ
Comments (32)
should I use refiner w this
Preview images didn't use it
Refiner wasnt used with the Sample Images. I only used a HiRes Fix
@KandooAI oh. in that case what HiRes Fix upscaler should I use? Also, forgive me, but one more question. when you HiRes Fix'd the preview images did you 2x? or another value. Thanks for any advice and thank you for this model!!
@mumbles I used the 4x_NMKD-Siax_200k Upscaler for those Images :)
I just wanted to give some props for the fast paced version releases. Keep it up!
Thanks for saying that :) Means a lot to me :) I guess the fast paced releases will slow down a bit in a month. But right now there is still a lot of stuff to fix. After that i will slow down the Updates on the Base Model and switch my Focus on Juggernaut LoRA´s ^^
Will you make an inpainting version?
Non of the SDXL models have an inpainting version it seems, as of right now.
I spent the night comparing V3 and V4 and concluded that V4 is more realistic, but not as creative and aesthetically pleasing as V3. Unfortunately, I don't like the V4 version.
Thats fair to say :)
Version 5 wont take that long.
For Version 4 the Focus was all on NSFW Data, so it shifted the Focus. It was needed to get the Human Anatomy stable (Female is fine, Male still 50/50) . Next Update will have a Focus on Backgrounds/Landscapes and Lighting and will prob switch back a bit.
Like i said in the Description, i´ll take you on the whole journey and be transparent the whole way. That also means that not everyone will like every checkpoint Version.
Just give me some time. Juggernaut on SD 1.5 was a total mess until Juggernaut Version 1.9 :D
Same.
It worsen all my Lora fidelity too
EDIT v4.5 fixes the issue AND shows improvements, well done!
This model was always my go-to and almost did the same prompt, but this v4 version is an 80% downgrade. I can't even do my usual prompts. It seems like this model is more focused on commers once again.... Sorry, but I am very disappointed with this model, as if this is not a real juggernaut model.
What's a "commer?"
I´ll see and hear your Critics since the Release of Version 4.
I´ll had plans to take the weekend off, but i am trying to get you a Fix ASAP. Can´t promise it today, but hopefully a Version 4.5 will be ready in the next 1-2 Days
I think there are some issues with v4 as well, but take your time. It's not like this is your job, right?
Please do take a break if you need one and look after your health and wellbeing! I'd hate to have another great checkpoint lost to burnout, we already have Lykon on indefinite leave from Dreamshaper. I can't speak for anyone else, but I'm happy to wait, no rush ❤️
Juggernaut is currently the best out there from my perspective, nothing I can't generate currently with it.
So do take you time! Life always comes before hobby ❤️
you rock. take your time
Hello, I found two major problems with v4. First problem is image quality, for some reason it is much grainier and jpeg-y than v3 and something off about the lighting. Second problem is its too horny... even with 10+ anti-nude prompts in the negative, I still frequently get mangled genitals when I want a SFW picture. Anyway please take your time and thanks for all your work. Juggernaut was my favorite photoreal model for SD 1.5.
Your models are wonderful. I really hope you continue with the Juggernaut XL line. But the v4 Juggernaut XL does look "overtrained". It is especially noticeable with non-photorealistic requests like "digital art". The results are "noisy".
Take as much time as you need. Your models are the best. At the moment I don't prefer V4, I'm still using V2 since it's the one I like the most but I keep donwloading and trying everything you release, I trust your creations will continue to improve! Thanks for the hard work!
Version 3 - the best XL model to date from all XL models I try on CivitAI! But Version 4, unfortunately much worse in my case. (I use Fooocus to generate images)
When you use version 4, you can use version 3 as refiner to dial in how much of the "realism" you want.
The speed of every release is crazy man, are you using a discord channel or something to communicate? I'm interested in helping with the development of it. Awesome work
Update:
After a long night, I've found the right checkpoint. To achieve this, I took an older training checkpoint from the current training set and merged it with Version 3. In the end, I tested 10 possible 4.5 versions and I think I've found the release candidate.
However, this also results in some changes compared to Version 4. The female anatomy remains stable. The male genitalia, on the other hand, has lost its "vigor". While it is still much better than in the previous versions, and especially the base model, the current model still produces more undesirable male genitalia than good ones.
Aesthetically, the look leans more towards Version 3, but it's of higher quality than Version 3. Midshots/Wideshots look better and after a HiRes fix, I've seen some really good images. Overall, Version 4.5 will appeal to those who enjoyed Version 3 :) Let's just chalk up Version 4 as a minor stumble. With Version 5, I'll collaborate with a small team of "beta testers" beforehand to avoid further mishaps ;)
Later today, I'll be uploading Version 4.5 (I still need to create the images for it :D). Thank you for your patience and the kind words I've received here and on Instagram :)
Appreciate you
Thank you for your hard work. I've tried some of your prompts in v4.5 and I have no idea why instead of getting "beautiful young girl" I would get a morbid and ugly girl instead LOL
😜😘
Oh, nice! I was testing V4 with some images I made on V3 and uhh... ya, I could tell something was off! Appreciate your hard work, can't wait to try 4.5!
I am very happy to continue testing for you, and I will provide feedback as soon as possible. Thank you for your work.
Haven't ventured into checkpoint creation and merging so I can only imagine how much work goes into it. Thanks for your efforts and continuing the journey - the progress of these models has been amazing. Please (please please!) consider revisiting male genitalia soon. Many of us are crying out for a model that can handle them.
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.