CivArchive
    IlluSatPony-Exp - IlluSatPony-Exp
    NSFW
    Preview 61735470
    Preview 61735479
    Preview 61735518
    Preview 61735561
    Preview 61735580
    Preview 61735606
    Preview 61735627
    Preview 61735638
    Preview 61735657
    Preview 61735674
    Preview 61735685
    Preview 61735754
    Preview 61735776
    Preview 61735785
    Preview 61735841
    Preview 61735867
    Preview 61735877
    Preview 61735901
    Preview 61735964

    What is this model? I have managed to mix Illustrious with Pony, after a long time experimenting I have finally achieved it, I decided to upload it in the Illustrious category since it is more loaded towards these models, as in most of my models what I seek is to get as close as possible to the realistic, what have I achieved by mixing these two types of very different models? First of all, something difficult to achieve is to get away from the typical Pony or Ill face, it greatly improves adherence to the prompt, there is a much greater variety of ethnicities and many variants that can be experimented with

    The parameters that I used are:

    • The VAE is not baked, use SDXL_vae

    • Upscaler: Im using Remacri_original upscale 1.5 Steps 10 denois 0.35 Hires CFG 6,

    • Width and height: 896*1152

    • CFG Scale: 6

    • Steps: 35

    • The sampling method i use is Euler A-SGM Uniform

    I added the EXP ending to make it clear that it is an experimental model so if you find any bugs help me improve for future updates, as I said in other models I like everyone to have access from the moment it is published, I don't like to put the models in early access, tips are appreciated and if you can't support me with a tip, I would love for you to share your images in the gallery, thank you for your support (use this model with care as it has a strong tendency to generate NSFW images)

    Description

    FAQ

    Comments (7)

    TootleMar 5, 2025· 1 reaction
    CivitAI

    Which checkpoints did you use? Cyber illustrious and real dream? I have been trying to do something similar, had the best results adding in the dmd2 Lora at ~0.5 strength as it boosts realism quite a lot.

    Sateluco
    Author
    Mar 6, 2025· 4 reactions

    I use my own models, SatPony and IllustriousSat-IOR, I added to the mix a lot of training steps and a little touch of sdxl and loras, when I start to experiment with base illustrious and pony I get an aberration mix, that's why I say it's not easy mix, I love DMD but you loose things like male generation, ethnicity and other things, thanks for the comments

    UrameshiMar 6, 2025

    @Sateluco looks pretty great, how did you manage to merge a pony model and and Illustrious model? I moved from pony to illustrious a few months back but I have great pony models Id love to merge with my illustrious models I just havent been able to make it work yet myself always comes out broken

    QH96Mar 6, 2025· 6 reactions
    CivitAI

    I think from all of the pony and illustrious models I've seen on this site you probably have the most detailed models. Every image is always packed with items, details, atmosphere and intricacies. Sadly, some prompt adherence is lost, but I don't know if there's any way that it could be maintained. Some good news for you; Civitai has announced that all checkpoints are going to be available for on-site generation soon. So your models will be getting a lot more attention.

    Sateluco
    Author
    Mar 6, 2025· 1 reaction

    Thanks for your comments and your feedback you always leave awesome images

    barinovrs433Mar 6, 2025· 1 reaction
    CivitAI

    Is it necessary to use score tags for this model as for Pony? Is there such a need?

    Sateluco
    Author
    Mar 6, 2025· 3 reactions

    You can use or not, all of my examples are without score tags but I saw there is a big difference when you use it, not better or worst, different

    Checkpoint
    Illustrious

    Details

    Downloads
    633
    Platform
    CivitAI
    Platform Status
    Available
    Created
    3/5/2025
    Updated
    5/17/2026
    Deleted
    -

    Available On (1 platform)

    Same model published on other platforms. May have additional downloads or version variants.