CivArchive
    Handcuffing (standing arrest) - Krea 2 - v1.0
    NSFW
    Preview 147589104
    Preview 147589106
    Preview 147589107
    Preview 147589109
    Preview 147589110
    Preview 147589111
    Preview 147589112
    Preview 147589113
    Preview 147589114
    Preview 147589115
    Preview 147589272
    Preview 147592816

    One woman handcuffing another while both stand: wrists pulled behind the back, one cuff locked and the other closing, the hands of the one doing the cuffing on the cuffs. Krea 2 on its own tends to pose the two side by side for the camera; this LoRA gets the actual interaction right. The roles go both ways: the officer can cuff the suspect, or anyone can cuff the officer. Trained on Krea 2 Raw.

    No trigger word. Prompt the scene the way it was captioned:

    • Name each woman by what she wears, then say who handcuffs whom in the active voice: A woman in a tan sheriff's uniform shirt handcuffs a woman in a denim jacket, standing behind her and closing polished steel chain handcuffs around her wrists behind her back.
    • For a role reversal, open with Role reversal, the security guard is the one being handcuffed. and then describe who cuffs her.
    • Cuff type: polished steel chain handcuffs or hinged steel handcuffs.
    • The stage is promptable too: one cuff locked and the second closing, both wrists cuffed, checking the cuffs.
    • Weight 0.8–1.0 (0.9 is a good default).
    • Strongest with both women standing and the cuffing behind the back. Scenes without uniforms work, but reroll a seed or two if the pose reads like a hug.

    Works for police, sheriffs, guards, bounty hunters, costume parties and comic-book art. All showcase images are made with this LoRA on a Krea 2 realism checkpoint, 8 steps, CFG 1.

    If you like my work, leave a comment, like, or whatever! It'll motivate me to make more and share more. Thank you!

    Description

    Rank 32 on Krea 2 Raw, 3000 steps (ai-toolkit, lr 1.5e-4, sigmoid timesteps). 92 crops at 512, of which the 44 sharpest also train at 1024 for cuff detail; 37 are close-ups of hands and cuffs and the 21 role-reversal images are weighted 2×. No trigger: detailed captions name both people by their clothes, who cuffs whom, the cuff type, the stage of the cuffing and every watch, bracelet and piece of belt hardware in view.

    FAQ

    LORA
    Krea 2

    Details

    Downloads
    280
    Platform
    CivitAI
    Platform Status
    Available
    Created
    11/1/2026
    Updated
    11/7/2026
    Deleted
    -

    Files

    hcstand2_krea2_000003000.safetensors