CivArchive
    Penis Arousal - v1.0
    NSFW

    Trained on the Wan 2.1 I2V 480P checkpoint using 20 5 second clips. It'll work best when the input image has a penis with clearly defined features.

    Trigger Phrase:

    penis becomes erect

    If you like this model and want to support the creation of future models, please support my work by leaving a positive review or posting your generations to the model page.

    Description

    FAQ

    Comments (25)

    GHDMRAug 6, 2025· 3 reactions
    CivitAI

    T2V version PLEASE

    BinaryBottleBakeAug 7, 2025· 1 reaction

    It most likely still works for T2V. I have used their Flaccid lora that says I2V on T2V and it works great!

    TompteAug 6, 2025· 1 reaction
    CivitAI

    🫡 🇺🇸

    Adaptalab0rAug 6, 2025
    CivitAI

    Glorious!

    itaskyAug 6, 2025
    CivitAI

    finally, thanks!

    whitespider9999Aug 6, 2025· 10 reactions
    CivitAI

    It would be amazing if you ever had the inspiration, to be able to make a lora similar to this, except with the erection growing as well. Since it makes less sense for the penis to remain the same size. Appreciate this regardless.

    jackaroo1432111Aug 7, 2025· 3 reactions

    If you just add "and larger" after "penis becomes erect" it works most of the time. I also use very simple straight forward prompts though, so the weight of the words really picks up well.

    whitespider9999Aug 7, 2025

    I'll try thanks.

    LocalOptima
    Author
    Aug 7, 2025

    The model actually does understand growing but I guess how much the growing appears in the final generation is kind of random and depends on the angle. The showcase doesn't do a great job of showing this. I'll upload some gens I made where it's more visible.

    bowiba1265909Aug 7, 2025

    Works well for me. And I actually like that it does not grow that much. Most of the penis loras go too unnatural big for my taste.

    whitespider9999Aug 7, 2025

    bowiba1265909 id just like the option for it to have a big shift from very small to very large. Anywhere between those two values can fit someone's personal preference for what they are going for.

    whitespider9999Aug 7, 2025· 1 reaction

    LocalOptima Thanks. I'm looking for some ways to generate from 100% flaccid to 100% erect. There's nobody that's really covered this yet. (besides possibly this) 

    HottieMiamiGuyAug 7, 2025· 7 reactions
    CivitAI

    As a gay man, I'm turned on by these, should suck her dry..lol great job!

    BigJuggsAIAug 10, 2025· 14 reactions
    CivitAI

    Fabulous work! Are you planning a version for Wan 2.2?

    LocalOptima
    Author
    Aug 12, 2025· 21 reactions

    Yes, I plan on retraining all of my loras on 2.2.

    BigJuggsAIAug 13, 2025· 1 reaction

    LocalOptima Can't wait to see it!!

    jackaroo1432111Aug 17, 2025· 8 reactions

    LocalOptima exciting! Do you think you could also do a "penis reveal" one? The old one on here got deleted and i loved that one for wan 1.2 😭

    WolphtizandoNov 28, 2025

    It's working on my wan 2.2 workflow. XD

    elmejor9369Dec 19, 2025

    @Wolphtizando are yos sure? i've tried in 2.2 but the penis stays down... any tips?

    tupuSep 18, 2025
    CivitAI

    Hi there, could this work in a reverse way? any idea how to propmt it, if posible? thanks

    gachmuch4uOct 1, 2025· 3 reactions

    maybe after you generate, you reverse the video?

    johnnywebb521626Oct 21, 2025

    You can try flftv

    SolidBoldOct 18, 2025· 1 reaction
    CivitAI

    could this just work for an already erect penis to help enhance accurate penis details? Maybe with just "penis is erect" or "erect penis"?

    JellaiJan 3, 2026· 1 reaction
    CivitAI

    I know this works for low noise Wan 2.2, but does Wan 2.2 have another lora that does the same specifically for 2.2? It's not really working that well for me. Maybe because I'm not working in 480p.

    jackaroo1432111May 25, 2026· 1 reaction
    CivitAI

    please consider making on of these loras for LTX!! :)

    I've been thinking about trying my hand at one for ltx, but not really sure how to caption the dataset or the training params. Any pointers?