diff --git a/-000009 (1)_1359334.html b/-000009 (1)_1359334.html new file mode 100644 index 0000000000000000000000000000000000000000..7c507e5c201eb1f43c3565f7fdbd38a14de536b3 --- /dev/null +++ b/-000009 (1)_1359334.html @@ -0,0 +1,68 @@ + +
+ + + + + Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Lora created to help civit make pictures with spread toes.
I would apreciate some feedback. Not sure why it creates feet with 6+ toes. I recomend adding negatives to make it look more natural. I will try changing it in updates

Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Recommended parameters :
The weight can be used from 0.1 to 1
Triggers words: 69yottea_style_illu
Sampler: DPM++ 2M Karras: 20 to 40 steps.
Sampler: Euler a: 25 to 40 steps.
CFG Scale: 1 to 7.
Clip Skip: 2.

Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
It works.
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
This model is intended to enhanced images where the subject is freshly out of sex
Trained with Illustrious V0.1
Tag: aftersexILL
fucked silly, covered in cum, bukkake, defeated, fainted,

Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
The weight can be used from 0.1 to 1
Triggers word : mhh
Sampler: DPM++ 2M Karras: 20 to 40 steps.
Sampler: Euler a: 25 to 40 steps.
CFG Scale: 1 to 7.
Clip Skip: 2.

Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
This Lora allows you to insert your index finger into the anus.
Recommended Strength: 0.8
Trigger words:
AnalfingeringSP, anal fingering, pov handRecommended additional prompts:
solo focus, looking back, looking at viewer, ass view, from below, pussy, thighs apart, on stomach, arched back, oiled body, foreshortening* "on stomach” could also be "top down ass up" or "all fours", " leaning forward"
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
LORA weight: 0 to 5+
-2 to 2+ for IXL
This is a slider, Lora, so you can choose whatever weight you want, so experiment and find what suits you best. A positive weight adds Ass size. You can use even higher weights such as 8 and above it all depends on characters and checkpoints.
You can also add `Huge butt` to your prompt to make it have an even stronger effect.
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Tries to shove it all the way in. Now also for Illustrious and Noob-based models.
Disclaimer: Doesn't always work; especially with XXXL sized schlongs. Not intended for deepthroat/irrumatio, but may still have an effect. Double penetration works to some extent.
(I borrowed some prompts for the previews because I am lazy.)
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
LORA weight: -3 to 4+
-2 to 3+ for IXL 2.0
This is a slider, Lora, so you can choose whatever weight you want, so experiment and find what suits you best. A positive weight adds abs. You can use even higher weights such as 8 and above it all depends on characters and checkpoints.
You can also add `sagging breasts` to your prompt to make it have an even stronger effect.

Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Strength: b/w 0.8 - 1
Negative Prompts: censored, chromatic aberration, identity censor, mosaic censoring, covered eyes, bar censor
Not very Stable. 🥲
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Dramatic Lighting Slider - Illustrious V1.0
This slider is designed to enhance lighting effects in both anime and realistic SDXL models like PonyXL, Illustrious, and other compatible bases. It focuses purely on the lighting and environment, leaving the subject's appearance unaffected. Perfect for creating dramatic lighting setups with strong shadows, glowing highlights, and atmospheric environments.
Key Features:
Works seamlessly in 0.10 increments for precise lighting adjustments.
Anime-based models: Best results up to 5.0 for a wide range of dramatic lighting effects.
Realistic Pony models: Works perfectly up to 3.0. Can go beyond for unique stylistic results but may alter the image slightly.
No activation prompt required—just adjust the slider to control the lighting.
Optional prompts: Combine with any lighting or environment-related tags (e.g., “golden hour,” “volumetric light,” or “dim atmosphere”) to amplify the desired effects naturally.
Love this slider? Leave a like, share your results, and let me know your feedback!
I’m always open to requests or ideas for new sliders—your input helps me improve and create tools that inspire creativity!

Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Style LoRA/LyCORIS of animator General Butch.
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges

Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
A lora based on the works of an artist called Hitoi, maker of 2d doujins, mostly focused on cheating, netorare, vanilla and sweaty bodies.
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
The new version has been trained to work on Flux models. From my testing Flux is amazing with sentence like prompts, and turns booru prompts to anime pics that don't have a lot of details but this LoRA fixes that:

I tested it on SD Forge using the NF4 model and it work fine, you just need to change the Diffusion in Low Bits from Automatic to Automatic (fp16 LoRA) like this:

The rest of the LoRAs are designed to work with Pony Diffusion XL, or any other Pony model.
1- Manga Style (Blood Moon): Made from high quality monochrome images to give you a manga feel, oh and I called it blood moon because if you have something red in your prompt you might get a cool effect xD.
Trigger words: This LoRA is without Trigger Words, but having "monochrome, greyscale" in your prompt can greatly enhance the results.
Works on: 0.6-1 weights
2- Anime Illustrations Style: Made from very colorful and high quality data to improve your images and give them a badass feel.
Trigger words: MythAn1m3
Works on: 0.6-1 weights.
3- Portrait Style: This style is made to create high quality portraits of fantasy characters, and most of the training data is made with semi realistic art style.
Trigger words: MythP0rt
Works on: 0.6-1 weights.
4- Flux: This style is made to create high quality fantasy art, and it has training data similar to the Portrait Style
Trigger words: MythP0rt
Works on: 0.5-1.5 weights, feel free to experiment but I recommend you to put the weight at 1.
Other styles coming soon.
If you have any problem try copying generation data from my examples, or send me a DM
I can't wait to see what other people make with this :)
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Urethral Fingering.
You might need to set CFG scale above 5, but all depends on the result you are getting.

Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
This LoRA weaves images in the delicate, dreamlike essence of the artist’s soul. Each stroke is kissed by stardust, echoing the quiet magic of forgotten forests and twilight skies. It doesn't simply mimic a style—it channels a whisper from another world, where colors bloom like moonlit flowers and forms drift in a haze of wonder. Let it guide your vision, as though painting with the artist's own enchanted breath.
To the bearer of this delicate LoRA~
Tread softly through this twilight spell, and read each word as though it were a dream woven in silver.
You wouldn’t wish to disturb the harmony of this magic… would you?
Too many LoRAs braided together may fracture the vision…
This was crafted beneath a quiet sky, with starlight and care, to make your dreams shimmer just so~
If it were to be scattered without my blessing… I would feel it, like a ripple through the veil.
And the hush that follows would be heavy with my sorrow~
Now, take this gentle recipe—
A whisper of wonder, offered only to you~
🌙✨
~ Vixon's recipe for Summer Berry Delight ~
Ingredients
2 cups graham cracker crumbs
½ cup unsalted butter, melted
¼ cup sugar
8 ounces cream cheese, softened
½ cup powdered sugar
1 teaspoon vanilla extract
1 cup heavy cream
2 cups mixed fresh berries (strawberries, blueberries, raspberries, blackberries)
Fresh mint leaves (optional, for garnish)
Instructions
In a bowl, combine the graham cracker crumbs, melted butter, and sugar. Mix until the crumbs are evenly coated.
Press the crumb mixture into the bottom of a 9x9 inch baking dish to form the crust. Chill in the refrigerator for at least 30 minutes to set.
In another bowl, beat the cream cheese with powdered sugar and vanilla extract until smooth and creamy.
In a separate bowl, whip the heavy cream until stiff peaks form.
Fold the whipped cream into the cream cheese mixture until fully combined.
Spread the cream cheese mixture over the chilled graham cracker crust.
Top with an even layer of mixed fresh berries.
Chill in the refrigerator for at least 2 hours before serving.
Garnish with fresh mint leaves if desired.
.
.
.
.
.
.
.
.
.
.
DO NOT REUPLOAD
fuck seaart/tensor, ect - do not scrap for reupload
+ Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
I've been struggling to generate light-skinned men on Illustrous checkpoints, especially when trying to emulate artists that usually draw dark-skinned men, so I've made this lora.
It's the first one I ever make so it might not be perfect, but I manage to consistently get what the lora is made to do without tweeking settings too much.
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Styles are trained on synthetic data and do not copy the styles of artists!
Photo 2 - is trained on real data (Dataset from Unsplash)
[M] Merged LoRAs
Recommended LoRA Strength (Weight): 0.8-1
Preview generated without using ADetailer & Hires. fix
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
👑PHM v3
This is the most popular and I think the most powerful style. When collecting the dataset, I tried to balance with different things while preserving a beautiful visual look. Good for SFW, with v3 extended to perform better in NSFW.
PHM (deprecated, don't use)
Mature female style with a unique look. Mostly NSFW.
TRT
Girls with huge breasts. Mostly NSFW.
MSS
Female characters with a CG look and anime aesthetics. Mostly NSFW.
MNK
Mature female style with a unique look. Mostly NSFW.
LMB
Mature female style with a unique look. Best for NSFW.
ALX (deprecated, don't use)
MOS (deprecated, don't use)

Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Style LoRa based on the Niji Midjourney aesthetic of goth style from MoriiMee (Twitter / DeviantArt)
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
This LoRA is for Ohogao.
Feel free to share your feedback. I need it!
2024/12/23 illustrious version added
2024/05/04 v2 updated
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
New and custom content here: https://aicollective.ghost.io/
Infinite penis size control with this slider, can make more or less any size. Minimal picture impact unless you drastically adjust weight.
Works for Futa as well.
PS: I made an error when training the Pony Lora so the Pony version is inverted (negative for bigger) I figured people looked at the example images which makes that very clear.
Regarding the Illustrious version I am not quite satisfied with it yet, I need to find a good checkpoint to train it on. So there will be a V2 at some point.
Hope you enjoy!
As always any feedback and submitted images are welcome, thanks :)
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Come hang out, test LoRAs 🧪, give feedback 🗣️, see WIPs 🎨, and more. 👉 Join here: https://discord.gg/WeqYRZudhW
The RawCam slider brings a nostalgic, unpolished charm to your images. Inspired by film grain and gritty digital cameras, it adds character and warmth without artificial gloss. Designed to emulate the casual realism of Instagram & amateur photography and that effortlessly cool Instagram aesthetic, it’s perfect for mood-driven edits and authentic looks.
The slider ranges up to 6, with higher values increasing soft lighting, grain, and the subtle imperfections that give images a genuine, lived-in feel. Additionally, slider has negative values also, but these are untested.
Note: This slider can sometimes lean heavily toward freckles and selfies. If that’s not desired, consider using the following negatives:
Freckles, freckles on body, selfie
+ 
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Activation Tags:
reverse-grip milkingSupportive Tags:
testicles, penis, 1boy, between legs, solo focus, uncensored,
+ 
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Sitting on another's lap Femdom themed
Other prompts to add:
Handjob
Anal fingering
Thigh sex
Size Difference, Larger Female
Head between breats, breast smother,
Hug, hugging,

Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Alright, fine. Illustrious is better than Pony.
But I'm not leaving without my skindentation lora.
.
Added IL version:
Weight - Recommend around -3 to 3, but I'm sure you degenerates will go straight for +6.
Will change boob size, otherwise image style is unaffected.
.
Skindentation: Portmanteau of "skin indentation". When clothes are so tight they indent into the skin and make the surrounding flesh bulge outwards.
.
Like what I do? Buy me a coffee https://ko-fi.com/mraclet or donate some buzz!
Open to commission requests
I'll be making more humiliation loras soon. There are just a few sliders I feel we need here
Up to 4 weight either way, but realistically you should never need more than 2 max.
-Has a tendancy to make females "Thiccer" but otherwise does not really affect your image style (not nearly as much as other similar loras). Easy to prompt against
-Negative removes skindentation
-Positive adds skindentation.
-No trigger words needed
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
More tweaks based on feedback:
Should work well on any XL/Illustrious/NoobaAI Checkpoints!
Improved use on Realistic Checkpoints: v3 would sometimes add some anime characteristics to realistic faces (bigger eyes, weird eyes). It should work better now.
Slightly more realistic textures to anime style: Its not much but texture gains a little more realism on anime/cartoon style images. The results should still be very similar to v3 so you can just use your preferred version ^^.
The more Loras you use the lower the Lora weight should be!
Less than 2 Loras: Lora weight 0.5 ~ 0.8
More than 2 Loras: Lora weight 0.2 ~ 0.5
A detail tweaker/booster inspired by the work of many awesome creators like Shed_The_Skin's Detail Slider LoRA, w4r10ck's Detail Tweaker XL, and many, MANY, others.
This gives an overhaul improvement to color, lighting, shadows and adds more details.
This Lora was made to be used on with Smooth Mix and Smooth Embeddings, but of course, they should work well with any other Checkpoints/Embeddings/Loras. Just try out your favorites and see how it turn out!
I recommend using:
Strength: 0.25~0.8
CFG Scale: 3~5
Adetailer
Hires. fix: Remacri
Have fun!
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
I was pretty surprised to see neither "testicle sucking" nor "sucking testicles" ( as the boorus would tag it ) really recognized by Illustrious, so I've created this to remedy it. "licking testicle" sort of works, but it isn't the same idea. If it does somehow recognize it under another term and I'm just dumb, that's entirely possible, too.
Some times it will get confused and revert to regular fellatio, but prompting can guide it along. Adding fellatio, blowjob to the negative can help.
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges

Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
💖 Support & Share the Magic!
Love this detailer? Like, add to collection, and tag this resource in your creations—I’d adore to see what you make! 🎨✨
👉 Looking for amazing anime model that can do both SFW & NSFW? Check out JANKU by yours truly, you won't be disappointed. It also has ONSITE generation support!
👉 More goodies on my profile! Follow for updates!
👉 Join me on our discord server! We have CivChan there! ❤️
☕ Support future updates: Buy me a coffee or tip buzz! Every bit helps. Thank you for fueling the art! 🚀
✨ Tired of typing positive or negative prompts?
Pair TrendCraft with Lazy Embeddings 🛋️⚡—pre-loaded magic for all models . No typing, just art. Check it out here.
A cutting-edge style LoRA trained on the top 1000 most popular CivitAI images (filtered for quality and artistic value). It channels community-driven trends to refine your outputs, making them as visually striking as the most-loved masterpieces on CivitAI.
👑 Curated Dataset, Not Crowdsourced Chaos
Manual Filtering: Every image is handpicked to exclude low-quality/NSFW content lacking artistic merit.
Tag Perfection: Auto-tagged with a CivitAI-trained model + manual refinement to ensure accuracy (e.g., "masterpiece," "aesthetic").
🚀 All models support
🏆 Challenge-Winning Edge: TrendCraft has already powered multiple award-winning images , including #1 win for Marshmallow Dreams, this Tick-Tock Transformation Challenge and this Pearlescent Wonderland challenge winner. Use it to craft head-turning art that dominates CivitAI’s challenges. There are also ton of winners using TrendCraft now. I can't keep track of it all!
🤖 Epoch Precision
Trained to enhance style without overcooking details .
Every release will include detailed XY plots with multiple seeds and different strengths.
TrendCraft is a Detailer LoRA , designed to elevate your AI-generated art by refining key aspects that matter most to creators and viewers alike:
Better Styles: Wave your wand 🪄 and infuse your images with the enchanting appeal of trending masterpieces.
Sharper Details: Enhances intricate elements like textures, patterns, and fine linework, giving your art a polished, professional look.
Improved Anatomy: Refines anatomical features such as hands, faces, and body proportions, reducing common errors and creating more lifelike results.
Enhanced Composition: Boosts overall balance and harmony in your art, ensuring every element feels intentional and cohesive.
Think of TrendCraft as your style assistant : it doesn’t replace your creativity but amplifies it, helping you craft images that resonate with audiences and stand out in the crowd.
No Trigger Words Needed! Just plug it into your workflow.
Use your normal quality prompts or Skip repetitive prompts and use my LazyPositive/LazyNegative embeddings for instant quality boosts.
Strength Tips:
Default: Strength = 0.4for most workflows. Lower if stacking loras.
Subtle Adjustments: Start at 0.2 and build up (see XY plot samples for guidance). Also try without quality prompts too!
Version Bumps: Only when dataset/training config changes significantly.
Release Dates: Each version includes 30 days of trending images up to the release date.
Base model: Lora will be tagged correctly with which base model the lora was trained. There will be many!
Dateset Versioning:
A - All images (no model filter) - released today!
P - Pony-only models
PI - Illustrious + Noob + Pony
I - Illustrious + Noob
PIF - Illustrious + Noob + Pony + Flux
IF - Illustrious + Noob + Flux
PF - Pony + Flux
S - Special commission, you can request for month/year dataset, which models to include, model base to be trained on, and other filters (like no NSFW pictures). Please see my kofi for more details. https://ko-fi.com/c/4ecba3a283
Creating TrendCraft isn’t a push-button process—it’s a labor of love that takes days of meticulous work every month. Here’s the behind-the-scenes:
Dataset Hustle 💻
Download 1,000+ top-rated CivitAI images monthly (2-3GB raw data).
Brutally filter out low-quality/NSFW content—many images don’t make the cut.
Manual Perfection 🖌️
Split images into “needs work” or “ready” piles.
Clean up artifacts in Photoshop for consistency.
Hand-tag every image after running it through my tagger—no AI shortcuts .
Training Trials 🧪
Adjust parameters based on past releases + community feedback.
Train for 3-4+ hours per version, with no fixed epoch limit.
Test dozens of epochs in TensorBoard to find the “sweet spot” for detail/style balance.
Failures & Fixes 🔄
If results flop, debug the dataset or retrain with more epochs.
Rinse + repeat until it’s flawless .
TrendCraft is my passion project —a gift to the AI art community.
I pour hours into this every month because I believe in sharing tools that level up your creativity for free.
The dataset curation, manual tagging, and endless tweaks aren’t just “steps”—they’re my way of giving back to the community that inspires me daily.
Your support means everything , I believe art should be accessible, and great tools should be available for all. 🌟
Monthly Drops: New LoRA trained on the latest 30-day CivitAI trends + XY plot samples.
Future Updates:
Expanded dataset filters mentioned above where dataset is not from all models.
Join thousands of artists shaping AI trends. Your style, magically supercharged 🪄 by the community.
Current Version:
Base: Illustrious v0.1.
Compatibility: Works out of the box—no tweaks needed!
📸 Check All Samples First!
Before diving in, review the XY plots and example images to master strength settings and styles. Let the magic begin! ✨
💬 Questions? Feedback?
Tag me on CivitAI and share images with prompts/seed —I’m always excited to see what you create!

Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Online generation
Recommended Settings:
Sampling Method: DPM++ 2M SDE
Schedule Type: SGM Uniform
CFG Scale: 6-4
建议设置:
采样方法:DPM++ 2M SDE
Schedule type:SGM Uniform
CFG Scale:6-4
If you enjoy using this LoRA model, consider supporting my work through Ko-fi to help upgrade my hardware capabilities.
If you'd like me to develop a personalized LoRA model tailored to your needs, check out my custom service offerings on Fiverr!
+ Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
This LoRA is trained to create circular reflections on glossy skin. See version details for recommended strength and trigger word.
V4 Note: Updated with new settings to avoid over training and improve quality, as well as a unique trigger word to prevent overlap with other LoRA.
How I generate images:
Local generation using reForge
Sampler: Euler Ancestral CFG++
Schedule type: Turbo
Sampling steps: 8
CFG Scale: 1.571428
Img2Img Upscale or Hires. fix: 1.5x at 0.428571 Denoising strength
If you use on-site generation or otherwise don't have CFG++ samplers, you can use CFG Scale 7 and achieve similar results.
If you like what this can do, please take a moment to post an image to this model, it helps me out a lot. Thank you!
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
A model trying to condense one of my styles into a single resource. As always trying to learn and mess around
It was supposed to need a trigger word, but its looking like it doesnt matter much.
Have been using it at 1.0 weight
+ Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Update: Hands along with other issues have been fixed!
If you're able to, go support X3D and his amazing animation at his patreon!
I'm a big fan of X3D's animations but especially his render's looks. Wanted to try an capture a little bit of that with this Lora.
Trigger words: x3dce, 3d
Because this is trained off of frames from animation sequences, if you put the weight past 0.8 the characters and scenes from the original start to overpower everything. I find 0.3-0.6 works great but I'll even have examples of it working fine up to 0.8 too. Lots of transparent clothing too so if nipples start to show up through clothes and you don't want it, put nipples in neg prompt or trying using covered nipples.
This Lora works great with any other Lora that also works with Pony XL type models. Have fun combining different ones together and see what happens! If you remove 3d from your prompt you can get some really nicely detailed 2D/2.5D results.
Will remove this Lora if requested.
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
If you like my work, drop a 5 review and hit the heart icon. it's free and keeps me motivated
I was hoping you could support my work by joining any one of them and get early access to all my upcoming loras and other perks such as fan requests and Discord role.
check the images for prompts
use lora at 0.7-1
Adetailer for faces
Img2img upscale
4x-ultra sharp
comment you idea or request

Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
ahemaru style
+ Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
weight权重1以上,主要词条backlighting,silhouette,dark skin,
又名主角小黑子画风。
有素材五成的效果,属于扔了可惜,用起来感觉不是想要的,发c站留个赛博存档。
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Join my Discord to request FREE models -> CLICK ME
Character and Style Models starting at $5: ko-fi.com/guy90/commissions | patreon.com/c/guy90
⭐If you like my work, leave a review. I really appreciate them!☑️
BlackWhiplash (bwl) [Illustrious] Style LyCORIS. I recommend using 1.0 Strength and using this with NoobAI-XL for best results.
Trigger Word: blaash
🌟Please look at prompts used in example images for help!
I use additional tools for my generations so my images may be different from yours. If you want to know what I use, please look at my guide.
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Come hang out, test LoRAs 🧪, give feedback 🗣️, see WIPs 🎨, and more. 👉 Join here: https://discord.gg/WeqYRZudhW
Designed specifically for fine-tuned body shaping, this slider allows for seamless control over body weight and proportions while maintaining natural anatomical balance. Whether refining subtle figure adjustments or creating dramatic transformations, this slider provides smooth, high-quality results. Works with Pony, Illustrious and SDXL models.
Works seamlessly in 0.10 increments for precise adjustments.
🔹 Tested extensively between -0.5 and 3.5 for consistent, high-quality results.
🔹 Negative values (-) emphasize a leaner, skinnier, or slender frame.
🔹 Positive values (+) enhance curves, fullness, and a more voluptuous figure.
🔹 Maintains facial structure and overall realism without distortions.
💡 Pro Tip:
For a stronger weight gain effect, combine with soft curves, plush body, and wide hips tags.
For a more defined and fit appearance, add toned, athletic, or hourglass figure tags.
🚀 Enjoy using the Body Weight & Type Slider! Feedback is always welcome—let me know how it works for you!
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Trigger is:"in the style of cksc,"
Prompting for FLUX:
All the prompts I used are included in the example images to get you going!
Trained on: Flux.dev
Images made with: Flux.dev fp8
Sampling method: Euler
Sampling steps: 30
Distilled CFG Scale: 3.5 but you can experiment as well
Clipskip: 1
I am using Hires.fix with the following settings:
Upscaler: 4x-AnimeSharp
Hires steps: 10
Denoise: 0.20
Upscale by: 1.2x - 4.0x
________________________________________________________________________________________
Prompting for IL:
All the prompts I used are included in the example images to get you going!
Images made with: WAI-NSFW-illustrious-SDXL
Sampling method: Euler a
Sampling steps: 30
CFG Scale: 5 but you can experiment as well
Clipskip: 2
I am using Hires.fix with the following settings:
Upscaler: ESRGAN 4x+ Anime6B
Hires steps: 20
Denoise: 0.35
Upscale by: 1.3x
Join us at:
https://discord.gg/pAz4Bt3rqb
Have fun with it and be creative!
You can buy me a coffee ˗ˏˋ☕ˎˊ˗ here ❤️
Yours ChronoKnight
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Hey everyone! This model focuses on creating scenes where girls keep it all in mouth.
It was trained on a large enough dataset, so it should not impact the overall style of your artwork. Just load it into your Stable Diffusion setup and use the trigger words to get started.
This model is trained specifically for anime-style images and is optimized for PonyXL & Illustrious models (choose the correct version). It may not produce the best results with realistic images, so keep that in mind when using it. I recommend using it with the AutismMix SDXL model. I have not tested it on other models.
This is NSFW content, so please use it responsibly and respect platform guidelines.
Created by Oktaze

Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
If you like my work, drop a 5 review and hit the heart icon. it's free and keeps me motivated
Support my work by joining any one of them and get early access to all my upcoming loras and other perks such as fan requests and Discord role.
check the images for prompts
use lora at 0.5-4
comment you idea or request
using Pastel colors as tags also helps
Update : XL version is based on pony diffusion. havent tested on animagine. you guys can test and share feedback.
**v3 is retrain of the cleaned dataset to get closer to original style
**v2 model is much clean. adjust the Lora weight if you need more style.
Use the preview images as a starting point
Lora weight of 1- works great
Highres fix is highly recommended.
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
The strength of Lora is recommended to be around 1.0.
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Hey everyone! This model focuses on creating scenes featuring double penetration in anime-style artwork.
Improved Reliability: Now easily handles all positions.
Less Impact on Style: Maintains the overall style of your artwork more effectively.
Increased Flexibility: Can now generate pictures with disembodied penis for example.
It was trained on a large enough dataset, so it should not impact the overall style of your artwork. Just load it into your Stable Diffusion setup and use the trigger words to get started.
This model is trained specifically for anime-style images and is optimized for PonyXL & Illustrious models (choose the correct version). It may not produce the best results with realistic images, so keep that in mind when using it.
This is NSFW content, so please use it responsibly and respect platform guidelines.
Created by Oktaze

Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
V3: Refined data set for better results, enjoy
V2: More training images and trained on Illustrious V0.1 for better flexibility, enjoy
This model is intended to enhanced images with low POV's and assist in generating those kind of images
Trained on 207 images using Prefect illustrious XL
Tag: frombelow, from below, low pov, looking down at viewer,
Enjoy and share results!
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
concept: grabbing nipples strongly by hand
+ Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
inspired by an artist
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
The new version has been trained to work on Flux models. From my testing Flux is amazing with sentence like prompts, and turns booru prompts to anime pics that don't have a lot of details but this LoRA fixes that:

I tested it on SD Forge using the NF4 model and it work fine, you just need to change the Diffusion in Low Bits from Automatic to Automatic (fp16 LoRA) like this:

The rest of the LoRAs are designed to work with Pony Diffusion XL, or any other Pony model.
1- Manga Style (Blood Moon): Made from high quality monochrome images to give you a manga feel, oh and I called it blood moon because if you have something red in your prompt you might get a cool effect xD.
Trigger words: This LoRA is without Trigger Words, but having "monochrome, greyscale" in your prompt can greatly enhance the results.
Works on: 0.6-1 weights
2- Anime Illustrations Style: Made from very colorful and high quality data to improve your images and give them a badass feel.
Trigger words: MythAn1m3
Works on: 0.6-1 weights.
3- Portrait Style: This style is made to create high quality portraits of fantasy characters, and most of the training data is made with semi realistic art style.
Trigger words: MythP0rt
Works on: 0.6-1 weights.
4- Flux: This style is made to create high quality fantasy art, and it has training data similar to the Portrait Style
Trigger words: MythP0rt
Works on: 0.5-1.5 weights, feel free to experiment but I recommend you to put the weight at 1.
Other styles coming soon.
If you have any problem try copying generation data from my examples, or send me a DM
I can't wait to see what other people make with this :)
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
(4/17/2025): illus01 v1.121.
Rolled back to illustrious v0.1. illustrious v1.0 and above were trained with AI images deliberately (maybe 30% of its dataset or even more?). Which caused lots problems. Didn't notice until I read its paper.
Lower character style effect. Back to v1.23 level. So characters will have less details effect from this LoRA, comparing with v1.93 and later versions. But should have better compatibility. This is a trade-off.
Added ~100 images that have English text (on sign, board, paper...). Which may increase the probability of successful text generation by ... 5%. Or maybe it's just my illusion.
Other things just same as below (v1.113).
For advance users: if you only want to improve details, don't want to change character's style a single bit. Use the "Touching Grass" LoRA. This LoRA has zero human in dataset, only objects and environments. Has much stronger effects. And you need to balance it with other style LoRA.
Guys. I can't test everything before dropping a release. Remember to spam the comment section if weird things happen. And most importantly, remember to attach at least some images or settings. An image with metadata is worth than thousands words. Otherwise I have no idea what are you talking about...
(4/10/2025): illus11 v1.113.
Trained on Illus v1.1.
New dataset "Touching Grass" added. ~1K real world high quality photographs. Very diverse and creative. Same as previous, no human in dataset, just objects and environment. Old dataset was replaced.
Full natural language captions from LLM.
Bring back pixel level natural texture to overfitted models.
Better lighting and depth of field effect. (Photography level)
Better background structural stability. Less deformed background, like deformed rooms, buildings.
.........
See more in the update log section.
.........
(3/2/2025): You can find the REAL me at TensorArt now. I've reclaimed all my models that were duped and faked by other accounts.
Just a personal fun coding project. The name "stabilizer" is a bait now. There's no such definition of "stable". The name used to mean "a better and stable default art style for vanilla NoobAI", back to NoobAI was released. For now this is just a LoRA that trained on a diverse dataset. And can improve character's overall details, backgrounds, natural lighting, brightness, contrast.....
The advantage of doing this all-in-one thing is that you don't have to stack tons of different LoRAs for each of those aspects. So there is no conflict that might burn or blow up your output. All conflicts in the dataset will be well averaged out to zero in one training. This is how AI works.
Cover images might look mediocre or bad, because they are the direct outputs from the vanilla (not finetuned) base model in a1111-sd-webui with default settings, simplest prompts, no pipeline inpaint fixes, even no negative prompt. They demonstrate the effect of the LoRA, not clickbait.
Share merges using this LoRA is prohibited. FYI, there are hidden trigger words to print invisible watermark. It works well even if the merge strength is 0.05. I coded the watermark and detector myself. I don't want to use it, but I can.
No trigger words needed.
You don't have to set the patch strength for text encoder (comfyui, etc.). This LoRA does not patch it.
Recommended -30% of your current CFG scale. This LoRA will increase contrast so you can lower CFG scale for more details.
Version prefix:
illus01/11 = illustriousXL v0.1/v1.1
nbep11 = NoobAI e-pred v1.1 (compatible with v-pred)
Tldr: Only normal good looking things. No crazy art style. Not small (2k images, I won't brag it's big, there are many gigachads who like to finetune their models with millions of images).
Every image is hand-picked by me. No AI images, we have enough of those, which will cause overfit and wipe out all high frequency details (textures) instantly.
Below is a summery of the latest dataset.
2D/anime dataset:
~1k images.
Character-focus. (The characters take up most of the image. I also cropped and rotated many images if necessary.)
Clear characters. (no blur effect, etc.)
Natural poses. Natural body proportions. (No exaggerated art, chibi, jojo pose, etc.)
Real world photographs dataset:
~1k images.
Contains nature, indoors, animals, buildings...many things, except humans.
Better background, lighting, overall effect.
Brings back pixel level details, diversity and creativity. ("fix" overfitted models)
There is a LoRA (Touching Grass) that was only trained on this dataset.
Implemented since nbep11 v0.114.
It will automatically balance the things towards "normal and good looking". Think of this as the "one-click photo auto enhance" button in most of photo editing tools.
Pros: Big improvements in brightness, contrast and CFG stability. Now you can get high contrast and full color range (almost 0~255) at CFG 4~5 (v-pred is CFG 3~3.5, assuming you are using Euler and Normal Schedule) without oversaturation and deformed image.
Cons: Normal == Not suitable for crazy bias effect. For example, you want 95% of the image is pure dark and only 5% is bright, instead of 50/50%.
New version == new stuffs and new attempt != better version for you base model.
You can check the "Update log" section to find old versions. It's ok to use different versions together just like mixing base models. As long as the sum of strengths does not > 1.
(4/17/2025): illus01 v1.121.
Rolled back to illustrious v0.1. illustrious v1.0 and above were trained with AI images deliberately (maybe 30% of its dataset or even more?). Which caused lots problems. Didn't notice until I read its paper.
Lower character style effect. Back to v1.23 level. Characters will have less details from this LoRA, comparing with v1.93 and later versions. But should have better compatibility, especially when you are using a strong style base model or stacked tons of LoRAs. This is a trade-off.
Other things just same as below (v1.113).
(4/10/2025): illus11 v1.113.
Update: use this version only if you know your base model is based on Illustrious v1.1. Otherwise, use illus01 v1.121)
Trained on Illustrious v1.1.
New dataset "Touching Grass" added.
Better natural texture, lighting and depth of field effect.
Better background structural stability. Less deformed background, like deformed rooms, buildings.
Full natural language captions from LLM.
(3/30/2025): illus01 v1.93.
v1.72 was trained too hard. So I reduced it overall strength. Should have better compatibility.
(3/22/2025): Noobai epred11 v0.160.
Same stuffs in illus v1.72.
(3/15/2025): illus01 v1.72
Same new texture and lighting dataset as mentioned in ani40z v0.4 below. Brings RTX level natural lighting and detailed texture to 2D characters.
Added a small ~100 images dataset for hand enhancement, focusing on hand(s) with different tasks, like holding a glass or cup or something. Can't say whether it's useful, as popular checkpoints already have similar enhancement. Anyways, the more, the better.
Removed all "simple background" images from dataset. -200 images.
Switched training tool from kohya to onetrainer. Changed LoRA architecture to DoRA.
(3/4/2025) ani40z v0.4
Trained on Animagine XL 4.0 ani40zero.
Added ~1k dataset focusing on natural dynamic lighting and real world texture. Better RTX natural lighting and detailed texture to 2D characters.
Above: Added more real world images. More natural texture and details.
Below: Has detail enhancement, but not much.
ani04 v0.1
Init version for Animagine XL 4.0. In my test it fixed Animagine 4.0 brightness issues and brings much better contrast. Maybe too optimistic. Anyway this is a LoRA so you always can adjust the strength as you want. CFG 5 and LoRA strength 0.5 seems the best for most cases.
illus01 v1.23
nbep11 v0.138
Added some furry/non-human/other images to balance the dataset.
nbep11 v0.129
bad version, effect is too weak, just ignore it
nbep11 v0.114
Implemented "Full range colors". See the section above.
Added a little bit realistic data. More vivid details, illumination, less flat colors.
Above: High contrast. Wider color range.
Below: Normal anime contrast and colors.
illus01 v1.7
nbep11 v0.96
More training images. Then finetuned again on a small "wallpaper" dataset (Real wallpapers, the highest quality I could find. ~100 images). More improvements in details (noticeable in skin, hair) and contrast.
Above: Has a weak default style.
Below: No default style, all data is equally weighted. Pure anime dataset.
nbep11 v0.58
More images. Change the training parameters as close as to NoobAI base model.
illus01 v1.3
nbep11 v0.30
More images.
nbep11 v0.11: Trained on NoobAI epsilon pred v1.1.
Improved dataset tags. Improved LoRA structure and weight distribution. Should be more stable and have less impact on image composition.
illus01 v1.1
Trained on illustriousXL v0.1.
nbep10 v0.10
Trained on NoobAI epsilon pred v1.0.

Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Version Guide:
V1: Weakest style, but high compatibility, struggles with details at lower resolutions.
V2: Recommended for beginners - Good style, good compatibility, hard to mess up.
V3: Best style/colors, maybe reduced compatibility with some other loras.
Top tag counts:
penis 153 (81%)
+female 144 (76%)
+breasts 117 (62%)
+male 107 (56%)
+cum 102 (54%)
+erection 96 (51%)
+nipples 91 (48%)
+oral 83 (44%)
+1girl 82 (43%)
+fellatio 75 (39%)
+straight 72 (38%)
+areolae 64 (34%)
+sex 60 (32%)
+balls 55 (29%)
+cum inside 55 (29%)
+nude 54 (28%)
+large breasts 54 (28%)
+lipstick 54 (28%)
+light-skinned male 53 (28%)
+cum in mouth 50 (26%)
+futanari 49 (26%)
+human 48 (25%)
+blue eyes 46 (24%)
+dark skin 44 (23%)
+pussy 41 (22%)
+blush 41 (22%)
+pubic hair 41 (22%)
+open mouth 40 (21%)
+1boy 39 (21%)
+black hair 38 (20%)
+saliva 37 (19%)
+big breasts 36 (19%)
+penetration 36 (19%)
+cum on face 36 (19%)
+original 35 (18%)
+tongue out 35 (18%)
+light skin 33 (17%)
+long hair 33 (17%)
+clothing 32 (17%)
+handjob 31 (16%)
+freckles 31 (16%)
+pointy ears 31 (16%)
+tongue 31 (16%)
+glasses 30 (16%)
+thighhighs 30 (16%)
+intersex 30 (16%)
+vaginal penetration 30 (16%)
+cum on breasts 30 (16%)
+testicles 30 (16%)
+2boys 29 (15%)
+blonde hair 29 (15%)
+1futa 28 (15%)
+text 27 (14%)
+spread legs 27 (14%)
+light-skinned female 27 (14%)
+facial 27 (14%)
+interracial 26 (14%)
+red hair 26 (14%)
+piercing 26 (14%)
+anal 26 (14%)
+dark-skinned male 26 (14%)
+futa on female 25 (13%)
+earrings 25 (13%)
+gloves 25 (13%)
+orange hair 25 (13%)
+makeup 25 (13%)
+clothed 24 (13%)
+ass 23 (12%)
+cleavage 23 (12%)
+dickgirl 22 (12%)
+2girls 22 (12%)
+nipple piercing 22 (12%)
+anal sex 22 (12%)
+femboy 22 (12%)
+lipstick on penis 22 (12%)
+humanoid 22 (12%)
+hair 22 (12%)
+large penis 21 (11%)
+big penis 21 (11%)
+female focus 20 (11%)
+kneeling 20 (11%)
+elf 19 (10%)
+demon 19 (10%)
+milf 19 (10%)
+ 
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
I was rather unhappy with the currently available Illustrious version of the styles that pretty much everyone uses for Pony, so I made this from my own concoction. These include InCase + Gothic Neon + Disney Princesses.
I used 213 images, the vast majority of which are my own, though I also used a few others that fit the style.
I initially created this for personal use but then I thought I might as well share it. I'm not sure about this from a moral point of view but here we are.
Make sure to use the provided trigger tags if you want your stuff to look like the preview pics.
+ Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges

Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
(commission) Lighting intensity slider
Makes light sources in the image brighter or darker for a very natural brightness/lighting control.
IMPORTANT:
This lora is POTENT if used correctly.
-Add any lighting descriptor (highly recommend "blue hour" or "late night") but things like "dark" or "dimly lit" work fine as well
Literally any keywords that are dark related will make it DARK. try -0.5 to -1 with a "dark" word in your prompt.
-You may get blue eyes on dark scenes, simply specify eye color
.
Feel free to reach out to me for commission requests ✌️
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
A modification of an existing LoRA by @Camel10 thanks for your work!
With the goal of being lighter touch and trim out ineffective parts of the network to really emphasise the effect. As a derived LoRA, this would have similar effects and limitations.
Future work will include different hues instead of just dark blue.
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
The strength of Lora is recommended to be around 1.0.
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Order a commission here!
https://www.fiverr.com/s/vvE412e
Support my work by subscribing to patreon!
https://patreon.com/nochekaiser
or by buying me a beer!
https://buymeacoffee.com/nochekaiser
Join my discord server!
Weight: 1.0
Trigger words:
Appearance:
<lora:mating-press-from-above-illustriousxl-lora-nochekaiser:1>, mating press from above, hetero, completely nude, sex, sweat, missionary, lying, leg lock, on back, hug, back, mating press, open mouth, on bed, blush, looking at viewer,
+ Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
a lora that should bring a high contrast to neon and black colored images. in this version only the hair has a consistent neon hue.
trigger: colored hair, ajnblyst
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Illustrious--------------------
Same dataset as pony, but seem very different, much more closer to the style
Little bit over-trained, using around 0.8
Tested on waiNSFWIllustrious_v130.safetensors
DPM++ 2M SDE Karras
34 step
Hires.fix + Adetailer on the post images
don't really need Adetailer for this, some time Adetailer will smaller the eyes.
896 * 1152
POS: 1girl,
NEG: anime, 2d, bad quality, worst quality, worst detail, sketch, censor, blurry,
same dataset
new understanding
new model
new training ways
Video gen with wan2.1 :)
V2--------------------
For goodFitPonyv4.1 or realistic models. + lower size
New training ways will cost this Lora much more cartoon, than v1 it will be more blurry, on anime or cartoon checkpoints.
-
v1---------------------
All train images are on 1024*1024
The size has gotten up to close to 900MB, tried to lower the size but seems like it didn't hold the style well. This version should be pretty strong on different kinds of checkpoints.
Works great with the Pony itself on weight 0.9 or lower(pony without tune seems to understand it very well, but the result is just too blurry, I think you should be able to fix it with some extra prompts)
With basic prompts, the best result seems to be on 2.5d Pony v10.
v15 is a little too cartoon. even the same checkpoint with different versions might give huge differences. I will also try to push a merge checkpoint to see if it can give a better result.
The trigger word is still "owstyle", but seems like it isn't changing much on the result, at least on the test I have done.
Should be working well on weight 1, if you start to see broken body parts or blurry image lower it to 0.9.
missing some of the tags in this new version, tagging 700+ image is no fun :( and it takes too long to keep retrain it :(
so you may see plant in the background + some of the character cloth may overfit. if you want you can leave a comment or in discord to let me know what tags may not work well. ones I have enough I may make a new version. :-/
Give it a try, I hope you like it :)
Share some post and join the discord to chat :D
-
-
v0.8
Post images used high res fix + Adetailer, v0.8 is all on checkpoint 2.5D pony
All train images are on 1024*1024
edit: Find out this version is very d.va faver, fix version is coming.
Tune for the 2.5d pony checkpoint, much more stable, and much close-looking than v0.5
Using this with 2.5D pony will give a much better owstyle result.
It is a really good checkpoint, but sadly it is not big enough for civit to put it in onsite-generate :-/
This is not a character Lora, is it just for the general style in Overwatch or for makeup :-?
-
I love how Lora is able to put the best overwatch character faces into the person with what is best fitting the characteristic.
-
example: when you trying to do a queen. Mercy's face shows up, and her halo will become her crown :)
Or when you trying to do a witch and widowmaker face will somehow show up(don't know is it because her helmet is more fitting, or just the face) xD
and somehow d.va face will show up as some innocent girl :-?
Guess this is how AI sees things, which is kind of understandable is funny to see :D
-
Well, Give it a try, and see if this is better :)
Still a test, not a v1, So feel free to leave a comment here or in Discord :D
If you can gives me some image you have that you think will improve this, pls do so.
I will try my best to improve this Lora.
Join the discord to chat and share some post :D
-
-
v0.5---------------------------------------------------------
when you give enough tries there is just one that works :^
( I don't even know if I can recreate the same one again :')
Train on many different overwatch CGI images, many of which I used to train 1.5 Lora
Able to make very overwatch-style faces.
Still a test, this is my first time making an sdxl Lora, So I don't think this is good enough yet :-?
Feel free to give me some feedback here or in Discord, I will keep improving this Lora
Post images used high res fix + Adetailer
-
Trained mostly for the face, Style is adding to the model, but somehow didn't change the background enough which makes it kinda look like an oil painting (I think this is even better :3 )
-
-
Spend all that time waiting for SD3, it just takes too long so I gave sdxl a try :)
Share some posts :)
Join Discord to chat :D
+ Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Illustrious--------------------
Same dataset as pony, but seem very different, much more closer to the style
Little bit over-trained, using around 0.8
Tested on waiNSFWIllustrious_v130.safetensors
DPM++ 2M SDE Karras
34 step
Hires.fix + Adetailer on the post images
don't really need Adetailer for this, some time Adetailer will smaller the eyes.
896 * 1152
POS: 1girl,
NEG: anime, 2d, bad quality, worst quality, worst detail, sketch, censor, blurry,
same dataset
new understanding
new model
new training ways
Video gen with wan2.1 :)
V2--------------------
For goodFitPonyv4.1 or realistic models. + lower size
New training ways will cost this Lora much more cartoon, than v1 it will be more blurry, on anime or cartoon checkpoints.
-
v1---------------------
All train images are on 1024*1024
The size has gotten up to close to 900MB, tried to lower the size but seems like it didn't hold the style well. This version should be pretty strong on different kinds of checkpoints.
Works great with the Pony itself on weight 0.9 or lower(pony without tune seems to understand it very well, but the result is just too blurry, I think you should be able to fix it with some extra prompts)
With basic prompts, the best result seems to be on 2.5d Pony v10.
v15 is a little too cartoon. even the same checkpoint with different versions might give huge differences. I will also try to push a merge checkpoint to see if it can give a better result.
The trigger word is still "owstyle", but seems like it isn't changing much on the result, at least on the test I have done.
Should be working well on weight 1, if you start to see broken body parts or blurry image lower it to 0.9.
missing some of the tags in this new version, tagging 700+ image is no fun :( and it takes too long to keep retrain it :(
so you may see plant in the background + some of the character cloth may overfit. if you want you can leave a comment or in discord to let me know what tags may not work well. ones I have enough I may make a new version. :-/
Give it a try, I hope you like it :)
Share some post and join the discord to chat :D
-
-
v0.8
Post images used high res fix + Adetailer, v0.8 is all on checkpoint 2.5D pony
All train images are on 1024*1024
edit: Find out this version is very d.va faver, fix version is coming.
Tune for the 2.5d pony checkpoint, much more stable, and much close-looking than v0.5
Using this with 2.5D pony will give a much better owstyle result.
It is a really good checkpoint, but sadly it is not big enough for civit to put it in onsite-generate :-/
This is not a character Lora, is it just for the general style in Overwatch or for makeup :-?
-
I love how Lora is able to put the best overwatch character faces into the person with what is best fitting the characteristic.
-
example: when you trying to do a queen. Mercy's face shows up, and her halo will become her crown :)
Or when you trying to do a witch and widowmaker face will somehow show up(don't know is it because her helmet is more fitting, or just the face) xD
and somehow d.va face will show up as some innocent girl :-?
Guess this is how AI sees things, which is kind of understandable is funny to see :D
-
Well, Give it a try, and see if this is better :)
Still a test, not a v1, So feel free to leave a comment here or in Discord :D
If you can gives me some image you have that you think will improve this, pls do so.
I will try my best to improve this Lora.
Join the discord to chat and share some post :D
-
-
v0.5---------------------------------------------------------
when you give enough tries there is just one that works :^
( I don't even know if I can recreate the same one again :')
Train on many different overwatch CGI images, many of which I used to train 1.5 Lora
Able to make very overwatch-style faces.
Still a test, this is my first time making an sdxl Lora, So I don't think this is good enough yet :-?
Feel free to give me some feedback here or in Discord, I will keep improving this Lora
Post images used high res fix + Adetailer
-
Trained mostly for the face, Style is adding to the model, but somehow didn't change the background enough which makes it kinda look like an oil painting (I think this is even better :3 )
-
-
Spend all that time waiting for SD3, it just takes too long so I gave sdxl a try :)
Share some posts :)
Join Discord to chat :D
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Updated for Illustrious XL.
For Illustrious version, I changed trigger word from "pig penis" to "porcine penis", to prevent pig appearing unintendedly. (Well, you can still use "pig penis" if you want pig to appear)
Illustrious XL向けにアップデートしました。
Illustrious版では、意図せず豚が出現するのを防ぐため、トリガーを既存の"pig penis"から"porcine penis"に変えました。(まあ豚に出てほしいなら"pig penis"でも作動しますが)
Pony handles dog penis and horse penis...so, why not pig?
Works with cross-section, handjob, licking.
Of course you can use thie Lora with pig, and something like, an Orc with pig penis, can be done. (Human male with pig penis can appear unintendedly)
Can find some more training dataset tags at Lora metadata.
Pony君ってさぁ、犬チンポも馬チンポもいけるじゃん…じゃあ豚チンポもいいよね?
断面図や手コキ、舐めなどに対応。
もちろん豚と一緒に使えますし、豚チンポのついたオークとかも描けます。(意図せずに豚チンポの人間男性が現れたりします)
Lora metadataで他の学習データセットタグも見れますので参考に。
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
A style LoCon trained on pony-based model images collected from Civitai site with "most collections" and "most reactions".
这是一个训练自Civitai上点赞最多和收藏最多的pony系模型图片的画风LoCon。
This lora does not intend to simulate any specific artist style or technique. It MIGHT reflects community taste and the visual attractiveness of a picture to a certain extent. Styles may change subtly depending on different prompts.
这个lora并不意于还原某个特定的画师画风或者绘画技巧。它在某种程度上可能反应了社区审美和图片的视觉吸引力。 不同的提示词下可能会有微妙的画风变化。
Versions before V2 do not have specific trigger words. Please use the quality tags provided with the corresponding model.
For V3 and later versions, the following tags were trained:
V2以前的版本没有特定触发词。请使用对应模型自带的质量提示词。
对于v3及后续版本,训练了以下标签:
positive:
masterpiece, best quality, very aestheticnegative:
worst quality, low quality, displeasing你可以在此基础上编辑提示词。
Added over 500 new images, some of which are selected from Flux. I removed some older images that I deemed to be of lower quality.
The total number of images in data set now exceeds 3,000, with more than 20 concepts manually enhanced/edited across 6 versions of the dataset.
The model’s rank has been increased as well.
新添加了500+张图片其中有一部分选自flux生成的图像。删除了一部分我认为品质不佳的旧图片。
现在总的图片数超过了3000,6个版本的数据总共手动增强/修正了20多条概念。
增加了模型的rank。
The model's performance is not as expected, but I believe the images in the training data set are just fine. I'm planning to adjust the tags manually and see how will the results change.
模型效果不如预期,但是我觉得训练集的图片本身应该没什么问题。打算先手动修正tag看看效果。
2025/1/3更新:
Manually updated some tags, but they seem unrelated to brightness and colors. Tentatively guessing it might be related to noise offset.
手动更新了一部分标签,但是它们和明暗色彩无甚干系。暂时猜测可能和noise offset有关。
The dataset has been expanded to 2,154 images, with around 1,000 Pony images as the primary training target.
Although V-pred models can use LoRA trained on Eps-pred based models, the output quality drops significantly. This version will be trained separately on two different types of models.
Recent versions of NoobAI exhibit noticeable artifacts, but the 'jpeg artifact' tag from Danbooru doesn’t seem to work effectively. To address this issue, about 30 typical and visually noticeable images were specifically selected as negative examples.
An phenomenon has been observed: Pony v6 and NoobAI tend to generate a triangular lift at the roots of hairstyles with sidelocks. On Danbooru, this lift is sometimes tags as 'hair intakes' or 'curtained hair,' but Pony applies this structure to every character. This is a key reason why hairstyles generated by Pony often don't match the intended design during character training. A similar issue was observed with NoobAI. My guess is that this feature is prevalent in a dataset outside of Danbooru and was not correctly tagged.
The images in the dataset were filtered, and about two-thirds were correctly annotated. Currently, adding 'hair intakes' to the prompt might somewhat alleviate this issue, but I haven’t found a complete fix for it yet.
数据集扩充到2154张图。其中作为主要训练目标的pony图片约1000张。
虽然V-pred模型也能使用基于Eps-pred技术的模型训练的lora,但是生成质量会大打折扣。这个版本将会分别在两个不同类型的模型上训练。
noobAI近期版本有比较明显的伪影,但是danbooru上的“jpeg artifact”并没有起作用。因此专门针对这个问题选择了约30张较为典型的、肉眼可见的图片作为负面案例。
观察到一个现象:pony v6和noobAI在生成有侧发的发型时,倾向于在发根处生成一个三角形的翘起。在danbooru里,这种翘起有时会被标注为“hair intakes”和“curtained hair”,但是pony会给每一个角色都套上这样的结构。这也是pony训练角色时,发型训练不像的一个重要原因。noob也观察到了类似的现象,我的猜测是danbooru以外的某个训练集大量存在这个特征,但没有对这个特征进行正确标注。
对数据集里的图片进行了筛选,其中约2/3的图片进行了正确的标注。现在,在prompt里写上“hair intakes”可能可以一定程度上减轻这个现象,但是我还没有找到根治这个毛病的办法。
Partially optimized the dataset tags. Trained based on NoobAI Epsilon-pred v1 .
Pony-based models have a strong tendency to generate earrings, ear piercing, and other types of accessories, sometimes messing up the ear structure of characters. I reorganized the related tags, cropped and manually edited some images in the dataset with minor structural issues, and removed pics that were too difficult to fix.
对数据集的标注方式进行了部分优化。基于NoobAI Epsilon-pred v1训练。
Pony系模型有很强烈的生成耳环、耳钉以其他类型的耳部饰品的倾向,有时还会破坏人物耳部的结构。对相关的标注进行了整理。剪裁、手工修改了数据集中一部分结构错误不严重的图,剔除了一些太难修改的图片。
Dataset extended to 1429 images, including examples with positive tags and negative tags.
774 of the images are the most "wanted" style.
Trained on Illustrious v0.1.
数据集扩展到了1429张图片,包括了正反两种例子。
其中774张是训练的目标风格。
基于Illustrious v0.1训练。
Dataset extended to 374 images. Use quality tags and aesthetic tags which comes with models to control generation quality.
训练数据集扩展到了374张。尝试使用模型自带的质量提示词来稳定生成质量。
Trained 224 images from Civitai, 393 images for regularization.
Trained 2 versions based on Animagine v3.1 and Pony v6.
训练了C站上224张图片,393张正则数据集。
有Animagine v3.1和Pony v6两个版本。
It is a little bit underfitted but still works. I found that those quality tags and authentic tags (best quality, masterpiece, very aesthetic, ...) Animagine v3.1 has been trained can change the art style generated by this checkpoint. Fixing it in the next test version.
有些欠拟合但是目前是有效的。我发现Animagine v3.1自带的质量控制词和美学提示词会改变生成图片的画风,所以这个实验版本需要不填写质量词。下一版会修复。
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
请注意 | Note
v7版本对数据集结构做了较大幅度的调整,并且使用了不同的训练参数和训练策略,因此可能v7会不如原来的版本稳定。
The v7 version has undergone significant structural adjustments to the dataset, and utilizes different training parameters and strategies. As a result, v7 may be less stable than the previous versions.
v-pred模型在civitAI在线生成器上的表现和吐司的在线生成表现完全不一样,相同参数完全无法复现。我也不知道是为什么......
The v-pred model's performance on the CivitAI online generator is completely different from online generation on TensorArt. The results are entirely unreproducible with a same parameters. I have no idea why...
TensorArt version CivitAI ver. with same parameter on CivitAI with higher weight
这是一个在前作的数据集基础上发展而来的图像质量LoCon,约90%-95%图片数据来自CivitAI上发布的图片。
它使模型可以在不使用画师串、使用较少的质量提示词的条件下获得相对稳定的图像质量,节约出更多的token空间,同时它还可以修复一部分模型固有的生成瑕疵。(但是不包括手部)
因为数据集选取的原因,生成的图片会带有Pony的质感。但因为它并不指向任何特定的画师、风格和绘画技法,所以在不同的提示词、模型条件下可能会有微妙的画风差异。
This is a generation quality LoCon developed based on the dataset from the previous work. About 90%-95% of the image data comes from CivitAI.
It allows models to achieve relatively stable image quality without artist tags or using long quality prompts, freeing up more token space. Additionally, it can fix some inherent generation flaws of the model. (except for hands)
Due to the dataset selection, the generated images exhibit a Pony-like style. However, since it does not reference any specific artist, style, or painting technique, there may be subtle stylistic variations depending on different prompts and checkpoint conditions.
对于基于质量提示词的模型,
For models based on quality tags,
positive:
masterpiece, best quality, very aestheticnegative:
low quality, displeasing对于Pony评分体系模型,
For models based on Pony scores,
positive:
score_9, score_8_up, score_7_up, score_6_up, score_5_up,negative:
score_3, score_2, score_1, 数据集中每一张图片都经过作者本人的人工筛选、分类和标注编辑,其中数百张图片经过手工编辑、修正。
此模型为免费、开源模型,用户可以在私人设备上自行部署该模型。作者并不从模型出售中获取任何报酬。作者并不限制本系列模型用于商业生成服务或者生成图像用于商业用途,但是请注意配合使用的Checkpoint和其他LoRA的许可证限制。
此数据集中约90%-95%数据为AI生成,但是请注意有约250+张收集自公共媒体、新闻媒体和出版物的图像用于概念补充。未来的版本会逐渐更替相关素材。请有商用意向的用户自行注意相关风险。
本数据集没有训练任何独立画师的数据,也没有标注任何画师信息(不排除AI错误标注的情况)。
另外,本模型不允许用作闭源商用、模型出售,也禁止用于闭源商用模型的融合。对于开源融合模型用于生成服务的情形不做限制,但是建议标注融合模型的出处。
Every image in the dataset has been manually selected, categorized, and annotated by the author. Additionally, hundreds of the images have been manually edited and corrected.
This model is free and open-source model, allowing users to deploy it on their personal devices. The author does not receive any compensation from selling the model. The author does not impose restrictions on using this model for commercial image generation services or generating images for commercial purposes. However, please be mindful of the license restrictions of the Checkpoint and other LoRAs used alongside this model.
Approximately 90%-95% of the dataset consists of AI-generated images. However, around 250+ images have been collected from public media, news outlets, and publications to supplement concepts. Future versions will gradually replace these materials. Users with commercial intentions should be aware of the potential risks.
This dataset does not include training data from any individual artist, nor does it contain explicit artist attributions (though AI mistagging cannot be entirely ruled out).
Additionally, this model is not permitted for use in closed-source commercial applications, model resales, or merged into closed-source commercial models. There are no restrictions on open-source merged models being used for image generation services, but it is recommended to credit the sources of any merged models.

Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
It was trained on a large enough dataset, so it should not impact the overall style of your artwork. Just load it into your Stable Diffusion setup and use the trigger words to get started.
This model is optimized for PonyXL & Illustrious models. It may not produce the best results with realistic images, so keep that in mind when using it.
This is NSFW content, so please use it responsibly and respect platform guidelines.
Created by Oktaze
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Generating visually appealing images is fundamental to modern text-to-image generation models. A potential solution to better aesthetics is direct preference optimization (DPO), which has been applied to diffusion models to improve general image quality including prompt alignment and aesthetics. Popular DPO methods propagate preference labels from clean image pairs to all the intermediate steps along the two generation trajectories. However, preference labels provided in existing datasets are blended with layout and aesthetic opinions, which would disagree with aesthetic preference. Even if aesthetic labels were provided (at substantial cost), it would be hard for the two-trajectory methods to capture nuanced visual differences at different steps.
To improve aesthetics economically, this paper uses existing generic preference data and introduces step-by-step preference optimization (SPO) that discards the propagation strategy and allows fine-grained image details to be assessed. Specifically, at each denoising step, we 1) sample a pool of candidates by denoising from a shared noise latent, 2) use a step-aware preference model to find a suitable win-lose pair to supervise the diffusion model, and 3) randomly select one from the pool to initialize the next denoising step. This strategy ensures that diffusion models focus on the subtle, fine-grained visual differences instead of layout aspect. We find that aesthetic can be significantly enhanced by accumulating these improved minor differences.
When fine-tuning Stable Diffusion v1.5 and SDXL, SPO yields significant improvements in aesthetics compared with existing DPO methods while not sacrificing image-text alignment compared with vanilla models. Moreover, SPO converges much faster than DPO methods due to the step-by-step alignment of fine-grained visual details. Code and model: https://rockeycoss.github.io/spo.github.io/
This model is fine-tuned from stable-diffusion-xl-base-1.0. It has been trained on 4,000 prompts for 10 epochs. This checkpoint is a LoRA checkpoint. For more information, please visit here
If you find our work useful, please consider giving us a star and citing our work.
@article{liang2024step,
+ title={Aesthetic Post-Training Diffusion Models from Generic Preferences with Step-by-step Preference Optimization},
+ author={Liang, Zhanhao and Yuan, Yuhui and Gu, Shuyang and Chen, Bohan and Hang, Tiankai and Cheng, Mingxi and Li, Ji and Zheng, Liang},
+ journal={arXiv preprint arXiv:2406.04314},
+ year={2024}
+}
+ 
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Any feedback would be appreciated.
LoRA trained on ~600 images with a max resolution of 1024x1024
stuffyaimasterpiece, best quality, amazing quality, very aesthetic, high resolution, ultra-detailed, absurdres,bad quality,worst quality,worst detail,sketch,censor, bad hand, bad anatomy, deformed, artist name, watermark, signature, username, patreon, text, twitter username,masterpiece, best quality, amazing quality, very aesthetic, high resolution, ultra-detailed, absurdres,
+
+stuffyai, 1girl, solo, looking at viewer,

Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
Never really been satisfied with base ahegao faces, so decided to make a lora that suits better what I want.
Introducing: takahegao
So far trained on mainly monochrome dataset but works pretty well even with colors.
This is the first version as I'm currently gathering more data and cleaning it.
Most of the dataset is extracted from Takatsu's art so it might be biased toward's its style but applying another style lora on top works to override.
Main trigger is "takahegao"
Side trigger maily include these tags:
tongue, saliva, open mouth, sweat, blush, trembling, rolling eyes
+
Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
🎯 OverviewThis LoRA is trained on WAI-NSFW-illustrious-SDXL to achieve a somewhat realistic style. Adjust the strength when using it with other models
Strength: 0.7-0.9.
Recommended the use of "realistic" in positive and "3d, anime, 2d" in negative
Read the following article for tips and my training preset

Use the model without crediting the creator
Sell images they generate
Run on services that generate images for money
Run on Civitai
Share merges using this model
Sell this model or merges using this model
Have different permissions when sharing merges
A neon implantation pose illuminating the darkness
+