This model is recommended for use with Comfyui
compatibility with Reforge is not guaranteed.
RECOMMEND setting (NTR MIX Xiii)
{your Prompt}
masterpiece, best quality, amazing quality, very aesthetic, absurdres, newest,
you can add tag scenery, (volumetric lighting), To make the background more distinctive.Use dpm ++ 2m sgm uniform or karras (Sometimes euler is okay...?)
Step 25, cfg 5 (can up to 5.8 by looking at the situation)
If you not use LoRA, author tags, etc., up the step to 28 EXTREAM (very scenery) setting (NTR MIX Xii)
masterpiece, best quality, amazing quality, very aesthetic, absurdres, newest, (scenery, volumetric lighting),
{your Prompt}
masterpiece, best quality, amazing quality, very aesthetic, absurdres, newest, (scenery, volumetric lighting)Use dpm ++ 2m karras or sgm uniform
Step 28, cfg 5.8 (can take it down to 5 by looking at the situation)SOFT (char only, simple background) setting (NTR MIX Xii)
Quality tags (please put Bottom of the tag)
{your Prompt}
masterpiece, best quality, amazing quality, very aesthetic, absurdres, newest,Use dpm ++ 2m beta or karras or sgm uniform
Step 20, cfg 5 (up to 5.8 by looking at the situation)EXTREAM (very scenery) setting (NTR MIX Xi)
Use dpm ++ 2m SDE (recommend SGM UNIFORM or Karras or BETA, choose your own!)
Step 28, cfg 5.8 (can take it down to 5.5 by looking at the situation)this need Two Quality tags, bottom and top (refers to the beginning and end of the prompt.)
Quality tags (please put Bottom and Top of the tag)
masterpiece, best quality, amazing quality, very aesthetic, absurdres, newest, (scenery, volumetric lighting),
{your Prompt}
masterpiece, best quality, amazing quality, very aesthetic, absurdres, newest, (scenery, volumetric lighting)SOFT (char only, simple background) setting (NTR MIX Xi)
Use dpm ++ 2m karras
Step 28, cfg 5.8 (can take it down to 5 by looking at the situation)Quality tags (please put Bottom of the tag)
masterpiece, best quality, amazing quality, very aesthetic, absurdres, newest, volumetric lighting
{your Prompt}
masterpiece, best quality, amazing quality, very aesthetic, absurdres, newest, volumetric lightingNegative Tags (all setting)
lowres, (worst quality, bad quality:1.2), bad anatomy, sketch, jpeg artifacts, signature, watermark, old, oldest, censored, bar_censor, (pregnant), chibi, loli, simple background, Recommendation setting (under NTR MIX 777)
Use dpm ++ 2m Karras (in i2i, use BETA and cfg 5)
Step 28, cfg 7Quality tags (please put Bottom of the tag)
masterpiece, best quality, amazing quality, very aesthetic, absurdres, Negative tags
lowres, (worst quality, bad quality:1.2), bad anatomy, sketch, jpeg artifacts, signature, watermark, old, oldest, censored, bar_censor, (pregnant), chibi, loli, simple background, conjoined, futanari, Description
FAQ
Comments (37)
Can you make realistic version of this one?
Why is the quality tag also needed at the end of the prompt. I also noticed that you add 'volumetric lighting' to your end tag.
I don't know how it works exactly. But it's because I've noticed a difference in quality when I've printed the image multiple times, so if you don't experience that, you can just tag it once.
@tkio Thank you your answer. I believe repeating a tag gives it more emphasis It's possible you could get a similar results by sticking the quality tags in a bracket with a multiplier
(masterpiece, best quality, amazing quality, very aesthetic, absurdres, newest, scenery:1.2) probably needs experimentation.
Can i use any pony lora with this one? or which loras can i use ?
We can't guarantee that 3.5 and later will use LoRAs. We've seen some pony LoRAs work in 2.1#, but in general, Illustrious xl-specific LoRAs should work.
It would helpful to add "about this version" information for each of the newer models instead of having to search through the comments.
Great model.
This Model it's litteraly amazing, thank u <3
There's only one word that defines this model: PERFECTION!!!!. Actually, I'm not even sure if that's enough, it's simply incredible!
Hi, I don't understand what it is: sd (1.5/xl), pony or other but I want a Flux version because I like what I see so much.
Is it planned?
This is the best model on civit ai, hands own
That's a practical result, good job!
Pos: (masterpiece, best quality, absurdres) ...
Neg: (worst quality, low quality, bad anatomy) ...
The above prompts still produce decent quality output.
sad part is ppl cant use it onsite gen.
Great model! Can generate using danbooru tag format or even with natural language!
That's great! Are there any plans to update the V-Pred version?
It is impossible to stop generating images. Whatever you try, everything looks cool! This is definitely voodoo magic. I hope the author will not abandon the model and will continue to update it
Fantastic model! I have around 10 illustrious checkpoints and this one is the best for me!
Finally! An actual good illustrious model.
Hands down the best model I've ever use. SFW, NSFW, Actions, Memes, does everything I throw at it. I don't know what portions of your soul you traded to make this.
V4 just works! Handles much better than either noob or illustrious.
This model is amazing and has fast become one of my new favorites :) It follows prompts very well and always comes up with something awesome. The only caveat, though, is that I don't think it has the same artist knowledge as NoobAi or Illustrious. The artist styles look off or get overpowered by the default style of the model. Maybe there's a fix but Idk.
Is there a list of artists acknowledged by these models?
This model is amazing! Hope it can be kept updated!
I often get noise around the eyes, pussy, and nipples. The rest of the image turns out great! I have all the recommended settings (cfg 5.5, karras sampler, etc). Any advice to fix this without inpainting?
Do you use ADetailer? If not then try it out. Use the first tab for the face. If you get noise on the other body parts then try the second tab of ADetailer for body. Does a great work in sharpening and adding prompted detailes.
Also if you want to tweak some of your images.
Send them to the img2img tab and go wild on it :D
Increase your steps or change your sampler
@whity00 Adetailer helped a lot, thank you very much!
For some reason I prefer 3.5 over 4. I just can't pinpoint why.
same
it's nice but doesn't play well with chara or style loras and forces large heads, so it might be way overfit with it's own base style
@yourneighbourRiz do you know how to fix large heads?
@9868686 you can try resizing it a bit with prompts such as small head, but it's already pretty integrated. Probably the best way would be to merge some "mature female" datasets which tend to have smaller and narrower heads in proportion to bodies and offset the model a bit
what are the differences between 3.5 and 4?
Do you consider making an Inpainting version?
Nice model. Going to start using this as my default Illustrious model.
What are the best settings to upscale images? Best model, denoise? If, for example, I only want to upscale x1.5
Okay, I figured it out somewhat. I changed the sampler to 2M + Karras, and use x4 AnimeSharp 6B, upscale by 1.75 and with denoising 0.3. Other minor problems fix in i2i tab with inpainting and sketches
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.