CivArchive
    Vampire Fangs - v1.0
    NSFW
    Preview 7775201
    Preview 7775196
    Preview 7775195
    Preview 7775200
    Preview 7775194
    Preview 7775198

    With the 1.1 version, I'm really pretty happy with this.

    If you want just fangs, neg prompt "vampire." I had some success with "fangs visible over lower lip."

    Very often Occasionally there are double sets, assymetrical, or random fangs. This is mostly because the model seems to have been trained that fangs belong to animals: dogs, cats, etc. The shotgun approach is what I use; sometimes it's going to be hit and miss but 1.1 hits more than 1.0.

    Both meet the bar I set for release: gives vampires fangs, but also gives people fangs without other vampire characteristics (skin, eyes, clothes). I also gave an example for giving fangs to a werewolf transformation. 1.1 even lets you draw fangs in Cyber Reality 4.2! I recommend a "photorealism" prompt even though it's a realistic output as it helps break the strong realism bent of the checkpoint. You can also use a photorealistic/art checkpoint, then use a realistic checkpoint in Refiner or Hires Fix.

    Description

    Very often there are double sets of fangs or assymetrical fangs - I apologize, probably there is some inconsistency in the training set with which teeth become fanged.

    However, it meets the bar I set for release: gives vampires fangs, but also gives people fangs without other vampire characteristics (skin, eyes, clothes). I also gave an example for giving fangs to a werewolf transformation.

    Also, realistic checkpoints flat-out refuse to draw fangs. I recommend using a photorealistic/art checkpoint, then use a realistic checkpoint in Refiner or Hires Fix.

    FAQ

    LoCon
    SD 1.5

    Details

    Downloads
    192
    Platform
    CivitAI
    Platform Status
    Deleted
    Created
    3/11/2024
    Updated
    4/22/2026
    Deleted
    9/23/2025
    Trigger Words:
    fangs
    vampire

    Files

    VampireFangs_v1.safetensors

    Mirrors

    CivitAI (1 mirrors)
    TensorFiles (1 mirrors)