CivArchive
    Mike Maius - マイク・マイウス (Wistoria: Wand and Sword, 杖と剣のウィストリア, Tsue to Tsurugi no Wisutoria) SDXL LoRA - v1.0
    NSFW
    Preview 132746494
    Preview 132746581
    Preview 132746486
    Preview 132746583
    Preview 132746493
    Preview 132746487
    Preview 132746488
    Preview 132746484
    Preview 132746490
    Preview 132746492
    Preview 132746483
    Preview 132746489
    Preview 132746533

    My first LoRA!!!

    Character LoRA for Mike Maius from the anime Wistoria: Wand and Sword.

    Trigger word: mike_maius

    LoRA strength: 0.7-0.9 (0.8 recommended)

    Base model: WAI-Illustrious-SDXL v17 or compatible Illustrious models

    Resolution: 832x1216 (portrait) or 1024x1344

    CFG: 5-7 | Sampler: euler_ancestral | Steps: 20-30 | CLIP Skip: 2

    Mike is the hyperactive tournament announcer of the Grand Magic Festival. His design features brown spiky upward hair, red-brown eyes, a fang tooth, athletic V-shaped build (broad shoulders tapering to a narrow waist), grey vest with buckle trim, and a wand with purple magic glow that he uses for public speaking.

    He has an incredible range of anime expressions: confident smirk, comedic panic with blank white eyes, chevron squint laughing, star/symbol eyes, and dramatic shouting poses.

    Example prompt:

    masterpiece, best quality, amazing quality, mike_maius, 1boy, solo, brown_hair, spiky_hair, brown_eyes, athletic_build, broad_shoulders, narrow_waist, confident, smirk, fang, upper_body, dynamic_pose

    Also try asking for low angle and high angle shots to see what you get!

    Negative prompt:

    bad quality, worst quality, worst detail, sketch, censor, blurry, lowres, bad anatomy, bad hands

    Trained with kohya_ss sd-scripts on WAI-Illustrious-SDXL v17. LoRA dim=32, alpha=16, 10 epochs, 67 curated images + 173 regularization images.

    There's only a handful of clips of him available so getting diverse screencaps is tough. His look may not be completely accurate, but as I get better synthetic images of him I'll try another pass to see if we can solidify his look and behavior.

    Description

    Initial release. Trained on 67 curated anime screenshots and synthetic images with 173 regularization images. Best results with WAI-Illustrious-SDXL v17.

    LORA
    SDXL 1.0

    Details

    Downloads
    13
    Platform
    CivitAI
    Platform Status
    Available
    Created
    6/4/2026
    Updated
    6/29/2026
    Deleted
    -

    Files