What's new? What's changed?

Can't afford, see u 2 weeks later

And they ask buzz for that. Bs.

Модель конечно хорошая, наверное, но денег нет

I like the style of V2! If some rating tags like general could be added to separate sfw images from nsfw, that would be awesome! If I only put 1girl in the prompt, it will generate a busty girl with sweat and breath vapor by default.

good

Love it

Just my 2¢

V2:

You will get more busty/curvy female body (especially hips and thighs area. I tried to include petite body, small waist, small thighs, but does not change much.

Good

thanks

Ok

nice

Bro, I have been using your models. Thank you

Wow

I still prefer the look of 1.0 out of the other 2, it transfer lora detail and style a bit better. and clearer details like veins

plant milk usually takes forever for me personally so this will get much more usage when it's public. it's very nice thank you

niiiiice

cool

good

good

let's bring back 1.5

it's better than 2.0 imo

モデル制作者さん、もうすぐストレスで脱毛しそうですね…頑張ってください！ファイト！

After another disappointing release, 1.0 is still king. 2.0 for whatever reason wants to blow all character proportions overboard and make them extra curvy, even when they're characters that are not canonically thick/busty. Ofc, you can still achieve normal results with enough prompting, but there's no reason for such a strong inherent bias towards blown out body proportions.

good！

Thanks for updating this model :)

very good

I seem to have an issue with skin tones. Here on civitai I'm getting decent results, but if I download the model and run it on stable diffusion I can't get that glossy skin effect. Colors aren't vibrant enough. Does anybody know if you need extra prompts or loras?

good

Females are always very thicc. Sometimes it's good, sometimes it's not what I am aiming for. Any suggestions on getting petite body structure?

Finally Free!

Good

2.0 is super plastic-y compared to 1.0 Everything turns out super shiny gross looking, even when using an artist mix. I'll stick with 1.0 thanks.

Waw.. Amazing 😍

Has a real issue doing male anthro characters for some reason. Heavily male-weighted prompts still producing female characters.

v2.0 seems to be a small improvement over v1.0 in some places but not all

(I posted picture grid for comparison)

- I enjoy the lighting more, its softer and more present for characters that don't have as a big dataset

- The eyes and hair received a small improvement

- The bodies tend to be thicker but you can just prompt it to be petite

- The line art is thinner or more cartoony which I like

- Colors are a little lighter which I personally enjoy

v1.0 is still amazing so its all personal choice

somehow v2.0 is very good for hires-fix or on higher resolution. good details and much more stable than using normal resolution.

man ai is such a great tool. or for porn slop lol

Hands are better now, but colors have become paler. Backgrounds and environment contain fewer interesting details. I like v1 than v2.

Amazing!

great!

it is just insane how good this model is
tysm

My humble review using this model:
(Check my profile to see results made with this model)

This model handles complex prompts impressively well and delivers solid hand compositions.

It also excels at generating characters interacting naturally with objects — something many models struggle with.

I’ve published over 200 images using this checkpoint, and rarely had to rerun the same prompt to get the result I wanted. It’s that consistent.

For best results, I recommend generating a batch of 5 images per prompt and selecting your favorite.

Overall, it’s a well-rounded, dependable model. I definitely recommend giving it a try.

good

This model is an excellent choice for users with standard hardware.
It offers impressive quality and decent speed for its class.

This is the first model that I’ve found that I personally think has the potential to catch up to the Wai Illustrious model. It still has some issues, chiefly the fact that the base Illustrious style still forces its way through even while using style tags which gives images a strong glossiness. Besides that it is an excellent model, capable of handling close to anything

I really love this model, and it is basically my main one. The only small issue that it has is that it is way too biased toward the girls having massive boobs. With which I normally like. The problem is when you want to create an image of a girl with medium or small boobs. You almost have to fight the model to get it.

Unfortunately 2.0 makes them booties too damn thicc for some reason. It's a issue when making more petite characters. I think 1.0 handled lighting better too.

Will v1.0 be available again for on-site generation?☺

any idea when version 1.0 is back up?

Wow

love you~~~

BRING BACK V 1.0!

please bring back 1.0

WE NEED 1.0 BACK!✊

Being back 1.0 please

Please add 1.0 back again!

Nice

Love it.

IF YOU GUYS WANT 1.0 BACK YOU HAVE TO BID FOR IT. IT'S NOT UP TO THE CREATOR

There's no point in either 2 or 1.5, they are noticeably worse and 1.0 is goated

Nice)

Love it!

impossible to use on comfy ui it only generates black images

I'm using this with comfy. The eyes are always messed up. Someone has tips to get them better?

great

nice

thanks！

thanks for sharing！

Why is it so great baby,why,looking my eyes,tell me why baby why😁😁😁

Personally, I still find the 1.0 model unbeatable, the best of the 3, but I would have liked to see more characters added, for example Hinata from Naruto, who is not there and can only be added with a lora

like magic 10/10 A+

i'm having trouble using this model to make full body shots, the shoes are always cropped out, no matter how I empathize the image to be a full shot or to make the subject full

gread model

The model is really really great, but I don't really like or really hate the eyes, it's more like it's in western style and a bit mature/milf female. It's not good in case if I want to generate a young girl like Ellen Joe from ZZZ, it's always looking like a goth mommy. Anyone know how to fix this? Btw, I already use style lora, but it's not fixing the issue.

really wish we had this as a style lora

edit: by it, i mean prefect 1.0.

好