Would appreciate comments, posts, reviews, and requestsif you'd like to pay for a commission or just send a dollar or two -&gt; <a target="_blank" rel="ugc" href="https://www.paypal.com/ncp/payment/U32UKLQUV9796">PAYPAL </a>or<a target="_blank" rel="ugc" href="https://www.paypal.com/ncp/payment/U32UKLQUV9796"> </a><a target="_blank" rel="ugc" href="https://ko-fi.com/bloodsplash/commissions">Kofi</a><a target="_blank" rel="ugc" href="https://patreon.com/Bloodsplash?utm_medium=clipboard_copy&amp;utm_source=copyLink&amp;utm_campaign=creatorshare_creator&amp;utm_content=join_link">Patreon</a>The Model I used for the V2 images is here titled "bloodsplash's Mix" <a target="_blank" rel="ugc" href="https://civarchive.com/models/12742/mightycorneo-good-model-overall-but-also-good-vtuber-model">MightyCorneo (good model overall but, also good Vtuber Model) | Stable Diffusion Checkpoint | Civitai</a>If censorship bothers you. (V2 is Now Available) I will be making an uncensored version either as V1.5 or V2, but in the mean time just add censoring, censor, bar_censor, mosaic censor, and any other censorship tags to the negative prompt, and add uncensored early in your positive prompt, completely slipped my mind as censors dont bother me. A lot of the images were uncensored so it should work.A multi view paizuri Lora. Corneo's pov paizuri Ti and Akto's clothed_paizuri pose Lora are great, but I wanted more variety as I am a near mastophilliac (meaning I LIKE breasts). So, I trained this Lora after repeated failures (this one is no exception) I got one that works pretty well.<a target="_blank" rel="ugc" href="https://civarchive.com/models/4475/corneos-pov-paizuri-embedding">Corneo's POV Paizuri Embedding | Stable Diffusion TextualInversion | Civitai</a><a target="_blank" rel="ugc" href="https://civarchive.com/models/6290/clothedpaizuri-pose-lora">clothed_paizuri pose LoRA | Stable Diffusion LORA | Civitai</a>I haven't tested with the above models; they could help.this Lora can do pov, side, above, front( similar to side), paizuri+fellatio, paizuri+irrumatio, double paizuri, autopaizuri, ,perpindicular paizuri (his one sometimes doesn't work), and can give some decent fellatio without paizuri.I suggest picking one Tag and not mixing them, although you could get some nice results maybe, haven't tried it yet.I also reccomend the disembodied penis lora for some of them.<a target="_blank" rel="ugc" href="https://civarchive.com/models/10871/disembodied-penises">Disembodied Penises | Stable Diffusion LORA | Civitai</a>recomended weights 0.6 - 1.0Might need to put monochrome and/or greyscale in negative prompttags:PAIZURI: OppaiZuriPOV: pov.SIDE: from_side.FRONT: from_front.BELOW: from_below. (one of the rarer ones to get successful)ABOVE: from_above,PAIZUR+FELLATIO: OppaiZuri+fellatio (optional tongue, fellatio, oral).PAIZURI+IRRUMATIO: OppaiZuri+irrumatio, (optional but recommended close-up, pov, pov_hands, head_grab, fellatio, oral, irrumatio, cum_in_mouth)Edit about the Irrumatio and fellatio It seems just putting them and not the optional tags somtimes helps and sometimes doesn't, so experiment.DOUBLE PAIZURI: double OppaiZuri, double_paizuriAUTOPAIZURI: futanari giving herself a OppaiZuri, autopaizuri, (probably the hardest since it wasn't trained on it).PERPENDICULAR PAIZRURI: perpendicular OppaiZuri, perpendicular_paizuri (from_side or pov might help with this one).im very new to this, so if anyone could offer any advice on this would be nice.The image can look better or worse if you add OppaiZuri before the tag you select, look at the final 2 images to see what i mean.the width and height can vary but they will usually work with 512x512, 768x768, but most of them should work with any width or height, it seems more prompt and seed dependent.

(Outdated) PaizuriAllinOne (different angles and types)

should be more compatible, but may be worse in some regards, next update will probably use a new dataset. I'm working on a standalone paizuri+irrumaio/fellatio Lora but the results haven't been that great so far. same as the other versions using a lower resolution and then high res fix may work better, since the higher res the less it seems to stay coherent.

Files with SHA256: 17424009a862a624df79d0457d708dc2b23b132893acaf985a1bf3d0262eaab3