This is an experiment I've been working on for several months - and I finally hit on something that works! Based upon XenoGASM v2.1 (before it got pushed so hard toward "realism") then some of my standard mix of mayhem and madness - and the a few select LoRas strategically merged in we've got this initial release of "XenoPLASTY"Added: LCM Version. If you aren't sure what this is and don't want to be bothered, stick with version one. If you want to learn more about LCM and reducing the number of steps to create generations - <a rel="ugc" href="https://civarchive.com/articles/3297/getting-lcm-samplers-and-models-to-work">read my article here</a> about setting it up. <h2 id="heading-20">Unique Features</h2><ul><li>Breast Expansion Sequences without LoRa's or Controlnet! Thanks to @LeonTalon's <a target="_blank" rel="ugc" href="https://civarchive.com/models/105759/breast-expansion-or-concept">Breast Expansion LoRa</a>, I managed to get it merged in and work (with surprising consistency) without bleeding into other generations through connecting it to specific blocks and in increments rather than a simple merge. So... all the triggers from that should work - and generally work as they would if you use the LoRa - but with this merge, it should also connect to some other areas and make it better and/or more interesting - at least I hope so. (More on this later). Starter tips for BE: Check the sample images... wide images tend to do a really consistent "before and after" concept while tall images at a 3:4 ratio tend to do a 4 shot one - but the order of images is often incorrect. There may be a way to improve this with prompting. We'll have to see. Example Seed Prompt: (Breast Expansion:1.2), (BETaut:1.1), (BETorn:1.4), (BEsequence:1.5) small breasts, gigantic breasts, (before and after:1.3), multiple_views, </li><li>MC/Hypnosis - much of this is a carry over from GASM v2.1 but I did merge @TwoMoreTimes89's <a target="_blank" rel="ugc" href="https://civarchive.com/models/48013?modelVersionId=52617">Brainwash Machine LoRa</a> back in here at lowish levels. At high levels, it was corrupting things, but I think I found the sweet spot. You will likely need to add some emphasis to the main trigger to get it rolling, though... I usually start with ((BrnWshM)) or (BrnWshM:1.2) and go from there. </li><li>HyperFusion Body Controls: This has a bit of @throwawayjm's <a target="_blank" rel="ugc" href="https://civarchive.com/models/186849?modelVersionId=209793">HyperFusion checkpoint</a> merged in for the extra body control that provides. I had to go moderately low with the levels, so those triggers (at least the unique ones) probably need some extra weight to get them going, too. (Higher levels of mixing were causing issues with hands and warpy bodies, plus pushing things toward the Anime/Hentai that model was trained on... expected, I suppose, when trying to make bodies pliable to transform at will lol). </li><li>...and more: There are a lot of other things in here that should work, but I either haven't had time to test or I haven't quite landed on the right prompt/trigger sequences to make them happen. Please feel free to experiment and share your successes! We're also always playing on my <a target="_blank" rel="ugc" href="https://discord.gg/5yv4AwHtan">Xeno Engine Discord</a> - both in learning and trying to get things to perform new tricks and in sharing what we've already learned. So please drop by and join the fun! (I may also post an article or two on getting some of these features to work... so stay tuned).</li></ul><h2 id="heading-21">General Features</h2>This should (hopefully) do many of the things my other Portrait models do - a little whimsy, interesting clothing choices, fun (albeit, not super detailed) background scenery, and so on.It can be a bit hard to push this to other styles, but most will work with some weighting. I need to experiment more with this and hopefully improve upon it all with version 2.There are a lot of "hooks" to things which I'm pretty sure are "in here" somewhere, but haven't quite figured out how to prompt, yet. (I'm more of a tech geek than a prompt artist - so maybe you can help!) For example, it seems to understand futas based upon some test generations - but I haven't figured out the best prompt combination to make them look good. Other fetishy transformations (though, maybe not unaided sequence shots).LoRas and Embeds all seem to work well with this, but remember - if you're using one of the ones listed above, they might help, but you'll likely need a lower weight than normal because those concepts already exist in it.<hr />Want Tips, Tricks, Pix of Things We Can't Show on Civitai? Join the <a target="_blank" rel="ugc" href="https://discord.gg/5yv4AwHtan">XenoEngine Discord</a> to discuss tips and tricks, complimentary Lora's and Embeds, and to post (and see) images not allowed here on Civitai.

Files with SHA256: 302789f00f0c0e837c380c3fac888f19b9084c258b485ae22e16c8dba6d001d0