CivArchive
    Preview 47976057
    Preview 47978133
    Preview 47982446
    Preview 47982244
    Preview 47987201
    Preview 47978947
    Preview 47978397
    Preview 47980739
    Preview 47982791
    Preview 48150844
    Preview 48153697
    Preview 48154250
    Preview 48179058
    Preview 48185433
    Preview 48185381

    GENOVA APEX is an experimental model.

    I created by merging my previous Gene Quantum model https://civarchive.com/models/885573/flux-gene-beauty with DEDISTILLED models and my own trained LoRAs. The process involved multiple stages and is purely an experiment to achieve high detail and realism similar to DEDISTILLED in just 8 generation steps(+ 8 generation steps Hires. Fix) .

    Version (UNet) – This version requires Flux VAE, Clip-L, and T5XXL to work effectively with the Flux development model.

    Recommended Settings for Optimal Generation

    • Sampling Method and Schedule Type: Euler Beta

    • Sampling Steps: 8 steps

    • I use Hires. Fix because it provides significantly better results: 8 steps, upscale 1.2.

    • My configuration: RAM 64Gb, VRAM 16Gb

    • Workflow for ComfyUi:
      I use a straightforward workflow for image generation, available here: Flux Photos DEV Workflow

      However, it requires a lot of memory because the process includes nodes for detail enhancement and upscaling, which also deliver the best results.

    • Please note that the quality on Forge may be lower than on ComfyUI. Here are the parameters I used: last update Forge, steps 8, euler beta, Distilled CFG Scale 3.5, CFG Scale 1 and I recommend use Hires.Fix Upscaler Latent or other be better, 8 steps, Upscale by 1,2 (higher upscale factors give better results but are slower and require more VRAM/RAM), Denoising strength 0,7, Distilled CFG Scale 3.5, CFG Scale 1 .

    • My Prompt Writing Assistant: https://civarchive.com/articles/7868/gpts-flux-photo-prompt-genP.S. The results I observe during the modeling process are even better than those I obtain from the final output of this process. Therefore, I continue experimenting, as the reason for this discrepancy has not yet been found...

    Description

    This model is more accurate and has better results

    FAQ

    Comments (8)

    sevenof9247Dec 30, 2024· 3 reactions
    CivitAI

    all only for comfyui ... only 5% of users use that

    tell us how in a1111/forge/swarmui

    and first step 512pix and than hires ??? and why hires i dont like that since SD15 the model should do it....

    DNA_1_618
    Author
    Dec 30, 2024· 3 reactions

    I work with ComfyUI, which is a highly flexible tool for working with Flux models. (For SDXL, I also often used ComfyUI or Fooocus.) The quality provided by just the base model isn't sufficient for me; if it were, I would simply use it. To achieve better quality, I use HiRes Fix and other workflow nodes.

    In ComfyUI, nodes are used, while in other applications, plugins serve a similar purpose. If you want to improve something in the application you're using, you'll also need to find appropriate plugins for it. HiRes Fix enhances image quality by first upscaling the image and then adding details through reprocessing. If you don't like this tool, you can simply choose not to use it.

    For now, I achieve the best quality in ComfyUI, and it seems that I haven't come across anything better, even when just using Flux 1 D. If you find something better, feel free to use the application that works best for you.

    econ_vectorJan 1, 2025· 1 reaction

    For a1111/Forge/SwarmUI, the best bet is to hop into their Discord and ask the developers to add compatibility.

    Also, a good way to get around needing HiRes fix is to use OpenAI's Dall-E. It's configured so the model just does whatever you prompted when you hit the button.

    chollman82141Jan 4, 2025· 8 reactions

    Only people using ComfyUI matter. Learn to use the real big boy tools or stop complaining.

    colinw2292823Jan 25, 2025· 3 reactions

    "Only 5% of user use that"

    Me thinks you just made that up good sir.

    rob52840May 26, 2025

    @chollman82141 Really?... Maybe it just boils down to techies und creatives... but then what do I know

    (the two approaches really do appeal to different types of people, different motivations, etc. Absolutely nothing whatsoever to do with big boys (although on a light aside, do you know the difference between men and boys? - the price of their toys. Nothing to do with the matter at hand, just thought I'd throw in a bit of light humor). Have a nice day .-)

    chollman82141Jun 7, 2025

    @rob52840 Well, as someone with a $3000 EUC, a $2500 electric scooter, and a $1000 RC truck I have pretty expensive toys lol

    rob52840Jun 7, 2025

    @chollman82141 lol! (but don't get me started.....!)

    Checkpoint
    Flux.1 D

    Details

    Downloads
    497
    Platform
    CivitAI
    Platform Status
    Available
    Created
    12/27/2024
    Updated
    5/24/2026
    Deleted
    -

    Files

    genovaAPEX_plus.safetensors

    Mirrors

    HuggingFace (1 mirrors)
    CivitAI (1 mirrors)

    Available On (1 platform)

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