CivArchive
    Preview 94347879
    Preview 93916965
    Preview 99198114
    Preview 106811119

    --Anima
    Trained on anima3. Recommended strength: 0.4.

    Works best with block filtering — I created a ComfyUI nodepack for this: ComfyUI-LoRA-Block-Filter. (use no_style_leak preset). Examples generated alongside my other LoRA — Lenovo.

    Experimental version. Results may vary depending on prompt and block weight settings.

    --Qwen2512
    Trained with the best my technique. Best use with Samsung lora

    -- Z-Image (Asian Edition)
    I'm experimenting with a completely different training method here. While it's still a work in progress, feel free to give it a try.

    Regarding the examples: I used 40 steps to test the limits, but you can definitely achieve good results with fewer steps like 20 for example

    -- Chroma
    Use with weight around 0.6-0.8. Works great along with lenovo lora


    -- Z-Image
    Unfortunately at this moment only european girls, but this was just test, so wait for new versions. Highly recommend to use with my workflow https://civarchive.com/models/2190193?modelVersionId=2466066

    -- Qwen
    Alternative version gives more clean result with less noise, blur, grain. I recommend when blend with other loras to set weight around 0.6

    Just a beauty girls for qwen. Use it along with my Lenovo lora for Qwen
    You can take workflow here: https://huggingface.co/Danrisi/Lenovo_Qwen/resolve/main/Qwen_danrisi.json

    Description

    FAQ

    Comments (23)

    2770379Aug 13, 2025· 12 reactions
    CivitAI

    I hope you release this for WAN. Really cool aesthetic.

    Danrisi
    Author
    Aug 13, 2025· 4 reactions

    okay, if you liked my taste that much, I’ll make a WAN version too. I originally made this LoRA just for myself — to test with the Lenovo LoRA for Reddit.

    2770379Aug 13, 2025

    Danrisi Yeah, it's a killer style. I really appreciate it!

    fox23vang226Aug 13, 2025
    CivitAI

    I tried it, it didn't work for me, kept cropping out my models head and upper body.

    aspie_pyAug 13, 2025

    I can’t get it to even activate in my workflow.

    Danrisi
    Author
    Aug 13, 2025

    aspie_py did u update comfy? i got same issue until update comfy and every custom node

    Danrisi
    Author
    Aug 13, 2025

    weird. what weight do u use? cause i use around 0.6

    fox23vang226Aug 13, 2025

    Danrisi I tried every single weight, from 0.1 to 1, all bad, but when I get some time this week I will retest it with an updated comfyui. I usually keep my comfy updated everytime i log in.

    Danrisi
    Author
    Aug 13, 2025

    fox23vang226 hehe, i remember the time when flux released and i didn't update for comfy for the half of year and everything was perfect until i updated...
    anyway, did u use my workflow or your own?

    aspie_pyAug 13, 2025

    Danrisi I am starting to think I’m on the wrong FP, will try and change later.

    aspie_pyAug 13, 2025

    Danrisi  Ok it was the update, working great now. Thanks!

    jm112368767Aug 13, 2025· 2 reactions
    CivitAI

    Works great - definitely more realism than without. Thanks for this!

    wXw77Aug 13, 2025· 8 reactions
    CivitAI

    It works quite well for the Q3_K_M model with the 4-Step Lightning Lora. Images look way more realistic than the base modeel.

    By itself with Lenovo on the Q3_K_M model it causes strong noise, maybe due to the old digital photos in the dataset being noisy and the model not being able to denoise properly, and even 50 steps are not enough. So I would highly recommend the Lightning Lora in addition for weaker cards, otherwise the waiting time is extremely long for a very poor output.

    JBsharpSep 27, 2025

    Would it be possible for you to share your workflow??

    rocklawSep 1, 2025· 2 reactions
    CivitAI

    can you make any of these for sdxl?

    it would be so dope for some y2k themed pics!

    WhyNaNWhenNyanSep 1, 2025

    There are already plenty of models for it in SDXL

    rocklawSep 1, 2025· 1 reaction

    @WhyNaNWhenNyan oh really? can you point me to some that are similar to this style?

    MrSmith2025Oct 30, 2025

    @rocklaw You found one that is similar?

    lowkeylayersSep 3, 2025· 3 reactions
    CivitAI

    tyvm, works great. is there a flux version of this too?

    SwaggerBoiSep 4, 2025· 5 reactions
    CivitAI

    This Lora still surprising me. Borderline alien technology.

    DiabolicXSep 22, 2025· 3 reactions
    CivitAI

    First of all, thank you for sharing the model, this is a great model that can be used to create many different styles, not just porn images 🤔👍, in combination with other models it can give really great results.

    PsyMetricsOct 13, 2025
    CivitAI

    It work's on chroma????

    Danrisi
    Author
    Oct 14, 2025· 4 reactions

    Not yet. Atm trained only one test lora for Chroma