CivArchive
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined

    Doesn't require a trigger word, but adding them to the prompt will help sharpen the background and make it look like an actual iphone photo, otherwise it serves as an all purpose realism enhancer and will make all images more natural.

    For Flux, the trigger is "iphone photo".
    For SD3.5 Large it's "iph0ne", try lower weight around 0.5.. first attempt at sd3.5, it might be bad.

    Trained with Ostris's ai-toolkit: Github page

    SD3.5 Large V2 Test version comparison - FLUX V1 vs V2 comparisons below

    FLUX
    V1 is iPhone Photo 6-2000 /// V2 is iPhone Photo 7-15000

    Description

    FAQ

    Comments (7)

    BinaryBottleBakeOct 17, 2024· 2 reactions
    CivitAI

    Would you mind sharing your prompts and settings for the images you made with the Flux Fusion model? I'm using the model and the Lora together but no getting as good results as you.

    Anibaaal
    Author
    Oct 17, 2024· 2 reactions

    Hello! I added the prompt and seed to all the main images now. The images also have Comfy workflow embedded, you can drag and drop in comfy window to see the prompts, Civit won't recognize them-
    I used euler sampler, simple scheduler, 4 steps, 864x1152 dimensions.


    If you use Forge, to get better quality try KL Optimal or beta scheduler, with euler or Flux realistic sampler, and use the option with fp16 lora in low bit setting if you can.

    I don't use Forge but found those gave similar results to comfyui when testing it.
    Using the simple scheduler in Forge gives terrible results, not sure why there's such a big difference with Comfyui.

    BinaryBottleBakeOct 17, 2024

    @Anibaaal Ok. Yeah I'm using Forge with Euler/Flux Realistic and tried out Beta and KL Optimal, similar dimensions. Not sure why they look different.

    Anibaaal
    Author
    Oct 17, 2024· 1 reaction

    @BinaryBottleBake It will always look different between Forge and Comfy even with same seed and params, because they do some calculations differently, but there is also other option in Forge to get more similar outputs to Comfy, under:

    Settings > Stable Diffusion > Compatibility > ComfyUI option > Apply settings.

    This way they will be the most similar, not the same but close.

    dsmirnof462Oct 18, 2024· 2 reactions
    CivitAI

    👍🏻✨✨✨👍🏻

    LuexOct 23, 2024
    CivitAI

    what's AnibalXL.safetensors??? How can I get it

    Anibaaal
    Author
    Oct 23, 2024

    Hi, it's a lora for SDXL, it doesn't have any effect with flux. I must have forgotten to remove it or in some workflows I used it as "no lora" option.

    LORA
    Flux.1 D

    Details

    Downloads
    8,075
    Platform
    CivitAI
    Platform Status
    Available
    Created
    10/17/2024
    Updated
    5/28/2026
    Deleted
    -
    Trigger Words:
    iphone photo

    Files

    Available On (2 platforms)

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