CivArchive
    DragonHeart - v1.0
    Preview 28850181Preview 28870243Preview 28860744Preview 28860775Preview 28860777Preview 28860778Preview 28862032Preview 28861956Preview 28861957Preview 28861958Preview 28862932Preview 28862911Preview 28862913Preview 28862915Preview 28862914Preview 28862916Preview 28862922Preview 28862928Preview 28862930Preview 28862931

    -This Model is New currently being tested.

    Recommended starting point without a lora:

    Sampler: dpmpp_2m, (creativity with sharper visuals)

    Inference Steps: 25, 

    Guidance Scale: 7

    Sampler: Euler: (Portraits and pictures with softer visuals)

    Inference Steps: 25

    Guidance Scale: 7

    TIP for home/private use: may require more vram when mixing many lora's together, Sold State HD recommended, adjust your guidance as you add or remove lora. more lora's = less guidance.(normally) combining lora's is a fun and exciting way to make new pictures.

    This is a mix of my two previous projects sixdragonseatingcake mixed with Irisheavenlyblue-3heartschasingthedragon

    Description

    My two previous projects Mixed -VAE baked in - Only one file required.

    sixdragonseatingcake mixed with Irisheavenlyblue-3heartschasingthedragon

    Thank you for your Buzz Tips, It helps me learn and make new stuff.

    Have fun, Don't break the rules ;-)

    Checkpoint
    SDXL 1.0

    Details

    Downloads
    86
    Platform
    CivitAI
    Platform Status
    Available
    Created
    9/11/2024
    Updated
    9/30/2025
    Deleted
    -

    Files

    dragonheart_v10.safetensors

    Mirrors

    Huggingface (1 mirrors)
    CivitAI (1 mirrors)
    Other Platforms (TensorArt, SeaArt, etc.) (1 mirrors)