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
Base Model got 50k more Steps
Second Dataset trained with the help from Dreamlook.AI with 58k Steps
FAQ
Comments (21)
Knowing what the Juggernaut 1.5 model can do, I am SUPER excited to find out what these guys do with XL! I will be all over this model at every step of its development!
KandooAI,
version 2 now lost artistic style. Almost all of my art prompts now becoming photorealistic. Why so? I'm very disappointed and sad ((
Juggernaut (both the XL and the 1.5 version) has always been focused on photorealism. It's completely normal that during fine-tuning, the model neglects other areas (in which it isn't trained) and tries to depict them in the style it was trained in during fine-tuning (in this case, photorealism). Since the focus of Juggernaut XL will continue to be more towards photorealism and cinematic in the future, perhaps another fine-tuned model would be a better choice for you (e.g., Dreamshaper by Lykon). Or of course, you can continue to use version 1 ;)
@KandooAI but it makes no sense to loose art styles. Do you know that your model knows most of the artist unlike other models? It makes it unique. And now You said you want to make another 100500 "photorealistic" model? :((
previous versions are always available if you think the latest doesn't fit your need
other than that, it makes no sense to direct where the author want to direct the model to
Your models truly are juggernauts! Incredible!
This is going to sound a little bit odd for a request, but... could you please incorporate some not so perfect female bodies in your checkpoint? I don't seem to be able to find any "plump" or "curvy" women bodies in your model. I've tried the usual trigger words that I've found to work in other models: plump, curvy, overweight, fat. The only trigger that seems to work is "obese". And that takes the woman straight from perfect body to Jabba the Hut. Surely, there are some nice in-between images you could include in your training for those of us who like a little meat on our ladies. 😁
Yeah thats possible :) I put that on my Todo list :)
@KandooAI My man 👍🏽
@97Buckeye I am prob gonna do another Midsize Side Model for just Curvy/Chubby Bodytypes and add that to the Juggernaut Base. I can´t promise that it will be perfect in the next update, but i will adress that problem :) Thanks for pointing that out :)
@KandooAI a lora specifically trained for that can do the job right?
@vitokeorlini225 Yeah a LoRA would also do the job. But Body Types is something that should come straight out of the Base Model, at least in my Opinion. So different Body Types can be prompted in the future directly from the Base :)
@KandooAI Thank you! Love your model - adding curvy ladies would be very welcome!
did you try my favorite "chubby"?
Thank you for your hard work. ❤
V1.6 is out
the new DPM++ 3M SDE Karras show very well
below my imagepost, one with my lora (better faces)
https://civitai.com/models/119376?modelVersionId=142718
CFG 3 for less dynamic colores
thx for taking my prompt, so I have seen it somewhere else and modified a little ;)
what you think about my loras -> faces https://civitai.com/models/119376/better-faces-sdxl
doo you train similar?
What, if any, refiner model should I use with JuggXL?
I didnt use any Refiner. But of course you can use the regular Refiner
I have had no success with the refiner model, even on base SDXL. The author of Fooocus wrote a custom workflow that makes it better, which is now also in Automatic1111 and a custom ComfyUI node, but I am not hopeful. The refiner does strange things like put heavy makeup on all the women and adds too much contrast, like a smartphone camera. And the refiner destroys any anatomy that was censored in the base SDXL.
If you want to add extra details, check the ComfyUI github discussions on injecting perlin noise to the latent, it looks a very promising technique.
Great model! Thanks for sharing!
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.