CivArchive
    All-in-One FluxDev Workflow for ComfyUI - v0.2
    Preview 23594494
    Preview 23594515
    Preview 23594554
    Preview 23594558
    Preview 23594557

    See GitHub page for more details.

    Overview

    I've created an All-in-One FluxDev workflow in ComfyUI that combines various techniques for generating images with the FluxDev model, including img-to-img and text-to-img. This workflow can use LoRAs, ControlNets, enabling negative prompting with Ksampler, dynamic thresholding, inpainting, and more. Please note that this is not the "correct" way of using these techniques, but rather my personal interpretation based on the available information.

    Main Features

    • Switch between image-to-image and text-to-image generation

    • For text-to-image generation, choose from predefined SDXL resolution or use the Pixel Resolution Calculator node to create a resolution based on aspect ratio and megapixel via the switch

    • Load ControlNet models and LoRAs

    • Sampling with model sampling Flux and Sampler Custom Advance node, based on the original official demo workflow

    • sampling with dynamic thresholding and Ksampler Advance node, enabling positive and negative conditioning with FluxGuidance

    • Simple inpainting

    • High-res fix like iterative upscaling with Tiled Diffusion

    Description

    fix wrong notes connecting to florence2 node

    FAQ

    Comments (4)

    apo1sAug 10, 2024
    CivitAI

    Is there swtch button or manually bypass it?

    LING_APE
    Author
    Aug 10, 2024

    You can use rgthree's Fast Groups Bypass to quickly bypass and unbypass a group, or alternatively you can open Show Fast Toggle in Group Header in rgthree's preference to enable a small icon at the top right of your group to quickly bypass and unbypass it. I'll add it to the next version.

    apo1sAug 11, 2024

    @LING_APE Its so complicated I manually bypassed them except txt2img but it didnt work

    diaserikcg114Sep 8, 2024
    CivitAI

    I would have liked an explanation of the choices in the different modules because a few things leave me perplexed.

    Among them: the negative prompt: until now, either it was useless or degraded the image, or it increased the rendering time x2/x3, so you might as well solve the problem by inpainting.

    There is also this very strange text in the middle of your prompt

    I understand that this is your vision but if you don't explain your vision, we can't understand it wanting to follow you.

    Workflows
    Flux.1 D

    Details

    Downloads
    668
    Platform
    CivitAI
    Platform Status
    Available
    Created
    8/10/2024
    Updated
    6/6/2026
    Deleted
    -

    Files

    allInOneFluxdev_v02.zip

    Mirrors

    HuggingFace (1 mirrors)
    CivitAI (1 mirrors)