Update 19.06.2026. Lustify V9 (Zenith) IS OUT!
This is the absolute final version of Lustify built on the SDXL architecture. You can find the full changelog and a breakdown of the new features there. <----------
Recommended Workflow.
Alternative Access & Support: Outside of CivitAI's Early Access, you can get the model via my Boosty.
Just like last time, after the early access ends, there will be a 1-week period where the model is available for "on-site generation only." After that, it will become available for everyone to download.
Alternatively, you can make a direct Bitcoin (BTC) donation, just make sure to DM me after that: 1CoMZvrXtRXbz7BCzdvxifSjHqhaf1yUZs
LUSTIFY
Lustify is a photoreal checkpoint optimized for creating high-quality imagery of women in explicit/sexual scenarios. The model has been trained on a hybrid dataset, meaning it natively understands both Danbooru tags and natural language prompting.
Note: While it has a strong NSFW bias, if you guide it properly, it is fully capable of generating stunning SFW photography, objects, animals, and fantasy environments.
Crucial Points Before You Download:
NOT Pony or Illustrious-Based: This is a standard SDXL checkpoint. Do not use Pony/Illustrious LoRAs with it; they will not work correctly.
Distant Shots & Anatomy Limitations: Like 99% of SDXL models, distant shots can sometimes suffer from distorted faces or hands. This is an inherent limitation of the base technology (4 channel VAE), not the finetune. It is incredibly easy to fix using Inpainting or ADetailer. Please utilize these tools rather than reporting this as a bug.
Experimental Nature: My training workflow involves a lot of unconventional, high-risk experimentation. Because of this, some prompt combinations that worked perfectly on other checkpoints might behave differently here. If you encounter unexpected behavior, feel free to give feedback so I can look into it.
Recomended parameters:
Sampler: DPM++ 2M SDE/DPM++ 3M SDE,
Scheduler: Exponential/Karras
Steps: 30
Cfg: 2.5-4.5.
Highres.fix: upscale by 1.4-1.5, denoising ~0.4
For the distant shots of people you probably wanna use highres.fix and/or adetailer.
Also, use reasonable positive/negative prompts, shizoprompting does more harm than good (with any SDXL model).
Some tags that have a big visual impact with this model:
Camera type tags: "shot on Polaroid SX-70", "shot on Kodak Funsaver", "shot on GoPro Hero", "shot on Canon EOS 5D", "shot on Leica T".
Photography styles: "analog photo", "glamour photography", "street fashion photography", "candid photo", "amateur photo".
Lighting types: "cinematic lighting", "neon lighting", "soft lighting", "dramatic lighting", "low key lighting", "bright flash photography", "warm golden hour lighting", "radiant god rays".
Film types: "Ilford HP5 Plus", "Lomochrome color film", "Fujicolor Pro".
Photographers: Alessio Albi, Martin Schoeller, Miles Aldridge, Oleg Oprisco, Tim Walker.
Others: "film grain", "bokeh", "dreamy haze", "technicolor", "underexposed", "low quality", "lowres"
Alternatively, you can give a read to this wonderful prompting guide by PromptGeek. It has A LOT of useful tips I didn't know about even after years of using SD.
Description
1) This is my gratitude to everyone supporting me and my work. I truly appreciate you guys!
2) Besides training the model further, I've also made some adjustments - the colors aren't THAT dark by default now compared to V3.
3) In my tests, V4 is the best of all the versions. But that's when it comes to my usage - it may be different for you.
FAQ
Comments (42)
just like a wow
can you make a lightning version of the v. 4.0 please? thank you for the great work.
Thank you. Yes, I plan to make either a lightning, hyper or dmd2 version
@coyotte oh, hyper is even better than lightning. Thanks!
Okay it's out
@coyotte downloading as we speak. Many thanks.
Might be better than Flux for amateur photography. The grainy effect is good.
Oh, just saw your prompts that you include scores tags as well, gotta try, amazing model 👌
Score tags are only on the pics that were generated with this workflow https://civitai.com/articles/6651/refining-pony-models-with-sdxl15-for-full-realism-the-right-way
They don't do anything at all for this model alone and come from the pony part of workflow
@coyotte ohh lol, thanks!!, loved how you add photography terms on this model
I would say, v4 is the most realistic Model out there, even better than Flux. Great Work!
Gawd damn it, Civit! Why won't you make this on-site generator already!?
Would you make an inpainting version of 4.0?
Okay, will do
Two days ago you were about to drop everything and now you come up with an even more amazing model, man...
This is all thanks to you and others who have expressed support for me
V4 is REALLY good!
In the midst of "it's so over,” You found there was, within you, an invincible "we're so back"
Negativity can be helluva motivator... As well as the support I was given by y'all
Very good model... I love this model.
Is best than the v3
Amazing update! Really enjoy the tweaks you bring to each version. Thanks for the work! looking forward to seeing what you bring in future
Wow ! It responds so much better to basic prompts than before ! AND it seems to work better with basic inpainting than v3 did (though when I tried background swapping (using Fooocus inverted mask) it did make the new background super dark, but i'll just use v2 for background swapping). GREAT JOB all round !
I'm glad you continued your work. V4 is really great, seems to have an even better understanding of anatomy in interesting poses. Also, thank you for not using the early access feature so many creators are jumping to!
Thank you!
This tech is open source, it was given us for free, and it should stay that way. Charging for it, even if it's some imaginary currency (buzz), is just sad IMO. I get it, people want to be rewarded for their time, but they might as well find a day job 💁
You know, I don't usually make comments like that, but given recent developments, I understand that you are a person who really works at your best and who cares about your work, so I want, in my small way, to help you find the "gaps" in the checkpoint so if you wanted to one day focus on something in particular to improve, you would have a small base.
I don't know if you might find them useful or not, but from what I've seen, being a mostly NSFW-focused checkpoint, these are some things you struggle to understand (or don't understand at all):
- some much less "classic" positions such as "full nelson" cannot be understood.
- sperm management is a bit weak (especially on faces), you really have to work hard to get satisfactory results.
-the size of penises is almost always the same and random despite being specified in the prompt.
-there are many facial expressions unknown to the model, such as sadness, which could be strengthened (I think).
-even the hairstyles are very little "varied" and you often have to be satisfied with what he proposes (for example the model does not distinguish between blunt bangs and parted bangs, so when it reads the word "bangs" it automatically generates blunt bangs).
For now, apart from these small things, there doesn't seem to be much else and the model behaves perfectly, always. (great work).
I repeat, I know nothing about how to improve a model, so if I have written heresies, I authorize you to send me to hell 🙆♂️
Keep going!👍
PS: if anyone else wants to add some small details, please write it below, let's make this model the best around
Thanks a lot for such a detailed review!
So, a couple of things I want to explain.
1) Initially i intended this model to be just a "porn" model that would know the basic stuff and generate it without mutations. It only later in the development (around V2) I started caring about the SFW aspect of it, BUT it was already too late as the main base was established in V1. That's why you don't have that extra control on the hairstyle or facial expressions. I don't really know how to fix that at this point. The penis sizes - yeah... I didn't really focus on that, my bad.
2) Regarding the positions this model doesn't know - I encourage you to read my article and try this workflow. You won't regret it. It's also useful for so much more than just poses. https://civitai.com/articles/6651/refining-pony-models-with-sdxl15-for-full-realism-the-right-way
3) The V4 is supposed to be the last one for a while, as I want to focus on a different model I have in stash AND I want to create loras specifically for Lustify (what isn't possible if lustify is always changing). The full nelson lora can be the first one then ✌️
@coyotte i like how it sounds 👌
Amazing work! Is it difficult to make inpainting versions? Really looking forward to an inpainting model for v4!
Amazing work! Thank you so much for V4 I know it was a tough time for you and it's so good to see you bounce back with another first class model.
the best REALISM ever!
I have not come across a better model than this in SDXL.
Thanks for the great work!
Congrats on the new release!
Everything I loved about V1 being the best realistic/amateur version, is now back in V4 with even better quality. Niche opinion here, but the feet/soles and hand qualities of V1 is what was missing on v2 and v3, but you managed it again on V4. Finally a replacement worthy of my limited hard drive space! Absolutely love the amateur look and feel you can get if you prompt for it, just amazing realism.
Damn it coyotte, you've done it again ol' chap!! Take my buzz :')
Thank you so much!
I'm working on a lora for V4 that focuses on the amateur/leaked nudes style photos specifically (no leaked photos used in the training, it just the feel is similar).
Looks good! Any plans for a Lightning version of v4?
Yes. I'm testing Lightning, Hyper, DMD2 and TCD loras right now to find which one would work the best for the checkpoint.
Okay it's out
Just add the Lighning lora (in the prompt or with a lora node) & you got your Lightning version.
Details
Files
lustifySDXLNSFW_v40Alpha.safetensors
Mirrors
lustifySDXLNSFW_v40Alpha.safetensors
lustifySDXLNSFWSFW_v40.safetensors
lustifySDXLNSFW_v40Alpha.safetensors
lustifySDXLNSFW_v40.safetensors
lustifySDXLNSFW_v40Alpha.safetensors
lustifySDXLNSFWSFW_v40.safetensors
lustifySDXLNSFWSFW_v40.safetensors
lustifySDXLNSFW_v40Alpha.safetensors
lustifySDXLNSFW_v40Alpha.safetensors
lustifySDXLNSFWSFW_v40.safetensors
adawa.safetensors
lustifySDXLNSFWSFW_v40.safetensors
lustifySDXLNSFWSFW_v40.safetensors
lustifySDXLNSFWSFW_v40.safetensors
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.