HomeDashboardModelsWiro AppsAppsPricing
Blog
Documentation

Task History

No results

You don't have task yet.

Go to Models
Models
Task History

Dark_infinity /
Suzannes-XL-Mix

Suzannes-XL-Mix

Suzanne's XL Mix

276Runs
0Comments
Licence

Tell us about any details you want to generate

Specify things to not see in the output

Click To See More Samples
1744815617 Report This Model

Suzanne's XL Mix

Suzanne’s Mix is made because of a request from Suzanne. I thought it would be a fun challenge. It's not meant for anyone else, but you can download and use it anyway. It attempts to capture some of the composition and aesthetic of RealCartoon-XL by 7whitefire7 and enhance it with some of the more risqué qualities and capabilities of select models with mature and adult themes. It was much more difficult than I thought it would be, and the earlier versions were too much of a failure for me to post here; therefore, you’ll see that Suzanne's XL Mix starts with version 3

Version 8.0

A new attempt at an old project, I wanted to bring in some refinement and knowledge into Suzanne's model. Version 8.0 is an experiment, but so far, it seems capable of generating some striking images while keeping most of the knowledge of Version 7.0. The newer version generates much more realistic images but still can summon up some of the aesthetics from RealCartoon to add just a bit of surrealism to a scene.

Version 7.0

Version 7 is similar in aesthetics to Version 6 but fixes a few small issues with NSFW prompting. Overall, it's a better-quality version, as I've diversified the base by adding LEOSAM's HelloWorld XL version 5.0 to the mix. Prompting has been improved without any quality issues.

Version 6.0 Official

Version 6 has overall improved quality and now uses a very broad base of NSFW models and LoRAs rather than relying only on a key few. This increases the versatility of the model, but not all prompts will be interpreted the same as in previous versions. However, as its abilities have been increased, the pesky watermarking of NSFW images has been reduced, and now they only show up occasionally and can be better controlled with some negative prompting.

A more detailed and enhanced version of this release will be coming soon.

Version 5.1 More Details

Version 5.1 is a relatively minor update that brings increased quality to images, adding more and sharper details with better contrast and color depth – all with no Lora needed. The enhanced effects are the most dramatic in semi-realistic art styles, but they work for realist and hyperreal art as well. Because this model works with an increased amount of noise, you may need extra steps – I’ve increased from 30 to 40 with this version. Some samplers can’t handle the increased detail; if your image comes out “noisy” (looks like a mosaic overlay), try a different sampler. DPP++ 3M SDE Exp, DPM++ 2S a Karras, and DPM++ SDE all work very well.

See the examples below comparing version 5 and 5.1. Images generated with 5.1 have better light and deeper shadows to show off the contours of physical bodies, as well as enhanced facial details and more intricate variations in the subject’s clothing and background. Some generations may differ in image compositions when comparing versions 5 and 5.1.

V

Version 5.0

Version 5 is an attempt to use a realistic model base. It’s rebuilt from the start and no longer uses the Hentai ProtoGenesis as a base but includes a new similar HCNSFW model. I may post that model for others to use, but for the moment, I’ll keep it under wraps. That base was pre-merged with the NSFW POV All In One SDXL Lora and then mixed in with Suzanne’s Mix version 3 and RealCartoon version 6. Unlike previous versions, version 5 used very specific block weights in the merge to add more weight to the blocks that added the aesthetic and feel of RealCartoon XL with the blocks of the other models that had the most influence on the more mature themes of the other models.

In my opinion, Version 5 is much better than the previous two versions in almost every way – all accept the watermarks. They mostly only show up when you attempt to produce hardcore images, and you can slightly suppress them with negative prompting. However, if you have more complex prompts that only include a few specific NSFW keywords, then you can often get quite a few images without them. But if you are willing to inpaint, you should have some really good results.

Version 4.0

Version 3 turned out well, but it didn’t quite have the look I was going for when I set out on this quest. Since the first version of Suzanne’s mix already included RealCartoon-XL in its recipe, I decided to simplify version 4 by using it at a higher ratio instead of including RealCartoon directly. The Deep Dark Hentai version 2 mix was capable of producing some surreal-looking images and was great at handling complex prompts, so I decided to mix it in as well with the previous Hentai ProtoGenesis and Suzanne’s Mix version 1.

The resulting model is good but tends to create more anime-style images than the semi-realistic look I had aimed to produce. In my opinion, the model is better in general than version 3 but is further from the mix’s objective.

Version 3.0

Suzanne's XL Mix version 3 was an attempt to use my Hentai ProtoGenesis alpha model as a base for the model, with more realistic models mixed in to bring it to an almost realistic model that still carried some dreamlike qualities. In addition to that model, I also mixed in:

It produces good images with a surreal quality that almost touches realism, and it’s quite capable of generating a wide range of NSFW imagery; however, it doesn’t come close enough to RealCartoon’s control composition and other elements in order to produce amazing images. Though not quite there, it's still good enough to share.

Model Credits

The following models have been or are currently used in a version of Suzanne's XL Mix. Credit goes to the model creators.

Models

View All
Fast Inference

wiro/FLUX.2-dev

FLUX.2 [dev] is a 32 billion parameter rectified flow transformer capable of generating, editing and combining images based on text instructions.
3
Fast Inference

wiro/AvatarMotion with Caption

Generate avatars from photos and animate them into engaging videos in one seamless pipeline.
5
Text to Image

meituan-longcat/LongCat-Image

LongCat-Image is a 6B-parameter model built for high-quality image generation, delivering strong multilingual text rendering, realistic visuals, and efficient deployment.
0
Social Media & Viral

wiro/Song Frame

SongFrame places you into a cinematic world, pulls the soundtrack directly from your YouTube link, and fuses everything into a polished video — effortless, emotional, and instantly shareable.
7
Fast Inference

Tongyi-MAI/Z-Image-Turbo

Z-Image is a powerful and highly efficient image generation model.
0
Text to Speech

openbmb/VoxCPM

Tokenizer-Free TTS for Context-Aware Speech Generation and True-to-Life Voice Cloning
0
Text to Image

meituan-longcat/LongCat-Image-Edit

LongCat-Image-Edit, the image editing version of LongCat-Image.
0
Text to Image

ByteDance/seedream-v4-5

Access the Seedream 4.5 API for fast high resolution image generation and editing. Simple integration, clear pricing, and support for text to image, multi image input, and advanced creative workflows.
7