Brand new model trained using synthetically-generated datasets from publicly-available and privately-trained models, combined with tagging using multimodal LLMs. I kinda love this thing. I hope you love it too.
This model has been tuned for excellent composition, aesthetics, and "soul." Characters come alive, scenery is dynamic, and there are a /ton/ of well-supported art styles. If ZBase can't make it, no one can ;-)
Please use LoRAs as low ratios (0.4-0.6) to accomplish softcore and hardcore NSFW, as the model has been made "flaccid" for general audience use. Erotica and nudes are fully doable (and quite good) with the base model however, so enjoy :)
Description
Final version I'm making for SDXL, I need to focus on Pony-based and then Flux-based versions. This is probably the peak of what I can do for SDXL anyway. Has a ton of SD3 and Flux datasets shoved in there. It understands depth and layering /so/ much better than any other SDXL model I've found. I hope y'all like it.
FAQ
Comments (8)
Hmm, your version 3.0 behaves a bit like a Turbo, Lightning, Hyper model, but doesn't burn to oblivion with higher CFG or step-count. But, it's also super stiff... challenge accepted.
Stiff?
@dustydecapod I mean, it has a certain style that won't budge, like everything is pretty much always colorful, nice shadows, clean ... isn't bad thing and i guess it's your personal preference.
@TijuanaSlumlord Ah, ya, unfortunately you'll have to use style loras to get it to move much in any other direction. I had to sacrifice a fair bit of style flexibility in order to achieve some of the other goals I had with this release -- higher levels of detail and sharpness, faster render times, better consistency and prompt coherence, and a significantly better rendering of depth and use of depth in composition.
@dustydecapod Yeah, sounds about right. Merging to many or merging LoRAs with a high strength will eventually kill flexibility ( depending on how you do it ). But, it seams to do what you wanted it to do, so everything seems fine, hard to produce anything bad looking :)
The results are nice, but its randomly blurred. Like it doesn't know how to handle bokeh. Which makes it harder to use.
Shoot me an example
Can I use this for my NSFW games / Comics?
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.