CivArchive
    LikeEatingGlass Mix style | Illustrious // 僕のヒーローアカデミア MHA/BNHA Boku no Hero Academia - Experimental
    NSFW
    Preview 111758609
    Preview 111758606
    Preview 111758600
    Preview 111758608
    Preview 111758605
    Preview 111758603
    Preview 111758602
    Preview 111758599
    Preview 111758607
    Preview 111758592
    Preview 111758594
    Preview 111758604
    Preview 111758601
    Preview 111758597
    Preview 111758610
    Preview 111758593
    Preview 111758598
    Preview 111758595
    Preview 111758596

    Good evening, bringing out another one of my personal mixes. This LoRA is heavily based on the style of BNHA/MHA anime, and it's great to create an OC inspired by that world.

    Boa tarde/noite, trazendo mais um de meus mixes pessoais. Essa LoRa é baseado inteiramente no estilo do anime My Hero Academia ou BNHA/MHA, foi feita especificamente para a criação de OCs baseado no universo do anime.

    Recommendations

    Strength 1.0 (in case of the Experimental, if the style isn't really showing, try going higher), in this case you can also go to 1.9, and lower it for more flexibility with the checkpoint. Same goes for your cfg, anything below 4.5 is perfect or recommended. Also try not to use too many LoRAs together as it can diminish the quality of your image fairly quickly, or a high cfg will also do that.

    If you want to support me, you can either donate, drop a like or follow me for future updates, either one helps. Thank you so much, every single buzz or support that I receive is going straight into my LoRAs.

    Description

    I'm not happy with this version! I reworked the entire dataset in order to make it more flexible and add more flexibility to the camera angles aswell, since both previous versions struggled to make full body, etc.

    To be fair, this version is a bit odd to me, on one hand it gives out some amazing results, but only if you use very high strength, I'm not sure what could possibly be doing this, but I'd guess it's either the parameters I've used, or the dataset was way too large and it didn't pick up the style as fast as it needs, so I will definitely re-train this once I gather enough buzz again.

    FAQ

    LORA
    Illustrious

    Details

    Downloads
    235
    Platform
    CivitAI
    Platform Status
    Available
    Created
    11/27/2025
    Updated
    4/28/2026
    Deleted
    -
    Trigger Words:
    mmha2

    Files

    cbnhastyle.safetensors

    Mirrors

    Huggingface (1 mirrors)
    CivitAI (1 mirrors)