<h3 id="you-can-make-money-with-my-models.-sell-what-you-generate-use-it-in-paid-work-run-it-on-paid-services-merge-it-sell-the-merge-all-of-it-no-permission-needed.-credit-is-appreciated-but-never-required.-new-releases-might-have-a-short-early-access-window-behind-buzz-then-go-free-and-free-is-permanent.-once-a-model-is-free-i-will-never-re-paywall-it-or-pull-it-for-a-paid-platform-and-you're-welcome-to-mirror-archive-and-rehost-it-so-it-stays-free-no-matter-what-happens-to-any-one-site."><span style="color:rgb(46, 255, 53)"><strong>You can make money with my models.</strong> Sell what you generate, use it in paid work, run it on paid services, merge it, sell the merge all of it, no permission needed. Credit is appreciated but never required. New releases might have a short early access window behind Buzz, then go free and free is permanent. Once a model is free, I will never re-paywall it or pull it for a paid platform, and you're welcome to mirror, archive, and rehost it so it stays free no matter what happens to any one site.</span></h3><h3 id="feel-free-to-mirror-any-of-my-loras-to-huggingface"><span style="color:rgb(8, 255, 0)">Feel free to mirror any of my loras to </span><span style="color:rgb(255, 248, 54)">HuggingFace</span></h3><p></p><p></p><p></p><p></p><p></p><h2 id="h3-v2:-10eros-max-beta6-highly-recommended!-updated-to-a-new-dataset-with-much-more-data-and-a-higher-resolution.-thrusting-and-squeeze-should-be-better-than-v1.-for-sampler-i-use-euler-with-beta-and-20-steps-on-10eros-max-beta6-and-use-comfy-kitchen-attention.-sparse-attention-will-lower-the-quality-noticeably-same-with-the-turbo-loras.-audio-can-get-out-of-sync-sometimes-but-is-primarily-a-prompting-issue-when-it-happens.-prompts-are-provided-in-the-example-videos.-prompting-is-very-important-for-h3-bad-prompts-will-give-bad-results-on-h3."><span style="color:rgb(92, 214, 255)"><strong>H3 V2: 10Eros Max Beta6 Highly recommended! Updated to a new dataset with much more data and a higher resolution. Thrusting and squeeze should be better than V1. For sampler I use euler with beta and 20 steps on 10Eros Max Beta6 and use comfy kitchen attention. Sparse attention will lower the quality noticeably same with the turbo loras. Audio can get out of sync sometimes but is primarily a prompting issue when it happens. Prompts are provided in the example videos. Prompting is very important for H3 bad prompts will give bad results on H3.</strong></span></h2><h2 id="h3-v1:-h3-version-works-on-both-fl2va-and-ref2va-and-stacks-well-with-eros-max:-post-trained-with-goonrl-(i-named-it-that-because-its-funny-also-put-a-shitpost-in-the-lora-metadata)-my-rl-pass-built-on-diffusionopsd-(on-policy-self-distillation-arxiv-2608.24646)-run-in-rounds-on-the-model's-own-generations.-the-objective-is-motion-completion-and-coherence-there-is-no-aesthetic-scorer-no-learned-preference-model-and-no-human-ratings-anywhere-in-the-loop.-not-sure-if-i-will-release-the-rl-method-since-its-experimental-and-takes-a-fucking-shitton-of-time-and-experimenting-to-get-right-and-i-really-don't-have-the-time-to-maintain-a-musubi-tuner-fork-for-it.-i-will-post-better-gens-with-a-higher-resolution-later-they-take-forever-to-do-over-0.4mp-5sec.-not-trained-on-audio-so-audio-may-be-mid."><span style="color:rgb(105, 255, 150)">H3 V1: H3 version works on both fl2va and ref2va and stacks well with Eros MAX: post-trained with GoonRL (I named it that because its funny also put a shitpost in the lora metadata) </span><span style="color:rgb(79, 255, 111)">my RL pass built on DiffusionOPSD (on-policy self-distillation, arXiv 2608.24646), run in rounds on the model's own generations. The objective is motion completion and coherence; there is no aesthetic scorer, no learned preference model, and no human ratings anywhere in the loop</span>.<span style="color:rgb(105, 255, 150)"> Not sure if I will release the RL method since its experimental and takes a fucking shitton of time and experimenting to get right and I really don't have the time to maintain a musubi tuner fork for it. I will post better gens with a higher resolution later they take forever to do over 0.4mp 5sec. Not trained on audio so audio may be mid.</span></h2><h2 id="ltx-versions-info-below"><u>LTX versions Info below</u></h2><p></p><h2 id="new-version-trained-on-sulphur-2-base!-it-adds-squeezing-the-breasts-around-the-penis-and-thrusting-plus-normal-titfuck!-prompting-should-go-similar-to:-a-2d-animation-shows-a-anthro-woman-in-front-of-a-man.-he-gathers-her-breasts-around-his-large-penis.-he-is-thrusting-his-penis-between-her-breasts-at-a-steady-pace.-she-is-looking-at-the-viewer-with-her-mouth-slightly-open."><span style="color:rgb(23, 255, 65)">New version trained on Sulphur 2 base! It adds squeezing the breasts around the penis and thrusting plus normal titfuck! Prompting should go similar to</span>: <span style="background-color:rgb(37, 38, 43);color:rgb(140, 143, 163);font-family:-apple-system, BlinkMacSystemFont, &quot;Segoe UI&quot;, Roboto, Helvetica, Arial, sans-serif, &quot;Apple Color Emoji&quot;, &quot;Segoe UI Emoji&quot;;font-size:14px"><strong>A 2d animation shows a anthro woman in front of a man. He gathers her breasts around his large penis. He is thrusting his penis between her breasts at a steady pace. She is looking at the viewer with her mouth slightly open.</strong></span></h2><h2 id="if-you-want-better-sound-you-probably-need-to-prompt-for-it.-it-should-also-generalize-and-work-on-non-furry-and-realistic-generations.-i-train-my-loras-to-generalize-and-to-have-minimal-style-bleed."><span style="color:rgb(13, 255, 56)">If you want better sound you probably need to prompt for it. It should also generalize and work on non furry and realistic generations. I train my loras to generalize and to have minimal style bleed.</span></h2><p></p><h1 id="old-info-below">OLD INFO BELOW</h1><p><span style="color:rgb(20, 250, 43)">Update for LTX 2.3. This version <strong><u>works much better </u></strong>than the LTX 2 version</span></p><p><span style="color:rgb(20, 250, 43)"><u>I use the ltx2 furry nsfw lora at a strength of 1 with this and distilled lora at a strength of 0.5</u></span></p><p><span style="color:rgb(20, 250, 43)">In the included videos<strong> (yes it works with anime and non furry the same as it does on furry.</strong>) there is the triple sampler workflow I highly recommend using I get much better motion using it vs the stock workflows. Idk if it works on t2v.</span></p><ul><li><p><span style="color:rgb(20, 250, 43)">LTX 2.3 needs detailed prompting for best results. Be specific about what you want and refer to the prompting guide LTX posted. The prompt I used for the included video is in the metadata just check the video for the prompt.</span></p></li></ul><p></p><p><strong>Update: Uploaded a better lora you should notice more stable motion</strong></p><p>A paizuri lora for LTX2. Took a while to dial in training settings but this one turned out pretty good.</p><p></p><h2 id="don't-use-the-distill-lora-at-negative-strength-on-ltx-2.3-advice-below-is-outdated-and-for-ltx-2"><u>Don't use the distill lora at negative strength on ltx 2.3 advice below is outdated and for ltx 2</u></h2><p>Workflow is embedded in the attached video. Works great on realistic and animated/anime.</p><p><strong>I use it with my motion lora </strong><a target="_blank" rel="ugc" href="https://civarchive.com/models/2344781/ltx2-better-nsfw-motion"><strong>https://civarchive.com/models/2344781/ltx2-better-nsfw-motion</strong></a><strong> and with the nsfw lora </strong><a target="_blank" rel="ugc" href="https://civarchive.com/models/2310920/ltx-2-i2v-nsfw-furry-multi-purpose-sex-lora"><strong>https://civarchive.com/models/2310920/ltx-2-i2v-nsfw-furry-multi-purpose-sex-lora</strong></a><strong> all loras at a strength of 1 The liquid LTX-2 Liquid physics lora was also accidentally toggled on for the furry gens but I don't think it had an influence on the output since there was no liquid in the video lol </strong><a target="_blank" rel="ugc" href="https://civarchive.com/models/2348942/ltx-2-liquid-physics"><strong>https://civarchive.com/models/2348942/ltx-2-liquid-physics</strong></a> I also use the distilled lora at a -0.4 strength with the distilled model. For whatever the hell reason in reduces any glitchyness I get in gens I have no idea why but I did multiple fixed seed tests with and without the lora. Someone smarter than me may know why but still no clue why. Idk if it works on t2v its made for i2v.</p><p>Audio can sound odd at times like with all ltx2 loras.</p><p></p><p></p><h2 id="disclaimer-and-terms-of-use"><span style="color:rgb(250, 82, 82)"><strong>Disclaimer &amp; Terms of Use</strong></span></h2><h2 id="this-model-is-provided-&quot;as-is&quot;-without-warranty-of-any-kind-express-or-implied-including-but-not-limited-to-the-warranties-of-merchantability-fitness-for-a-particular-purpose-and-noninfringement."><span style="color:rgb(250, 82, 82)"><strong>This model is provided "AS IS", without warranty of any kind, express or implied, including but not limited to the warranties of merchantability, fitness for a particular purpose, and noninfringement.</strong></span></h2><h2 id="in-no-event-shall-misticrain69-be-liable-for-any-claim-damages-or-other-liability-whether-in-an-action-of-contract-tort-or-otherwise-arising-from-out-of-or-in-connection-with-use-of-this-model-or-any-output-generated-with-it."><span style="color:rgb(250, 82, 82)"><strong>IN NO EVENT SHALL MisticRain69 BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH USE OF THIS MODEL OR ANY OUTPUT GENERATED WITH IT.</strong></span></h2><h2 id="18+-only.-you-the-user-are-solely-responsible-for-anything-you-generate-with-this-model-and-for-ensuring-your-use-complies-with-all-applicable-laws-and-the-civitai-terms-of-service."><span style="color:rgb(250, 82, 82)"><strong>18+ only. You, the user, are solely responsible for anything you generate with this model and for ensuring your use complies with all applicable laws and the CivitAI Terms of Service.</strong></span></h2><h2 id="you-may-not:"><span style="color:rgb(250, 82, 82)"><strong>You may not:</strong></span></h2><h3 id="produce-anything-illegal-non-consensual-defamatory-or-used-for-harassment-deception-or-fraud"><span style="color:rgb(250, 82, 82)"><strong>produce anything illegal, non-consensual, defamatory, or used for harassment, deception, or fraud;</strong></span></h3><h3 id="create-any-content-depicting-or-appearing-to-depict-minors-(real-or-fictional)-in-a-sexual-context-no-exceptions-ever"><span style="color:rgb(250, 82, 82)"><strong>create any content depicting, or appearing to depict, minors (real or fictional) in a sexual context — no exceptions, ever;</strong></span></h3><h3 id="use-this-model-to-depict-any-real-identifiable-person-celebrity-public-figure-or-private-individual-in-any-context-sfw-or-nsfw.-no-likeness-use-period."><span style="color:rgb(250, 82, 82)"><strong>use this model to depict any real, identifiable person — celebrity, public figure, or private individual — in any context, SFW or NSFW. No likeness use, period.</strong></span></h3><h2 id="if-you-post-outputs-to-the-gallery-that-violate-the-civitai-tos-you'll-be-blocked-and-reported.-don't-do-this-shit.-don't-be-fucking-sus."><span style="color:rgb(250, 82, 82)"><strong>If you post outputs to the gallery that violate the CivitAI TOS, you'll be blocked and reported. Don't do this shit. Don't be fucking sus.</strong></span></h2><p></p>

H3 - LTX 2.3 - Paizuri - Titfuck (With Thrusting + Squeeze) I2V

<p>Larger dataset with higher resolution training.</p>

H3_Paizuri+squeeze+thrust-V2.safetensors

Files with SHA256: 4d2d0acc7e093c08e1dd1442184e70aa54dedaf5c2f7a4197613607bb1ea36ef