ComfyUI Workflow for FLUX.1 Dev - Two-Pass Text-to-Image Generation

This advanced workflow for ComfyUI is specifically designed for FLUX.1 Dev, a text-to-image AI focused on generating high-quality images from textual prompts. This setup is structured to achieve detailed and refined results through a series of comprehensive processing steps.

Key Features:

1. Two-Pass Generation: The workflow applies a double-pass on each image to maximize realism and detail accuracy, producing a more polished and coherent outcome that aligns closely with the input prompt.

2. Face and Hand Correction: This pipeline includes correction modules to enhance facial features and hand details, two challenging areas for AI image generation.

3. Upscaler: An upscaling module is integrated to improve the final image resolution without compromising quality, making the results suitable for high-resolution displays or prints.

4. Refinement: The refinement stage further adjusts details for a more precise and visually pleasing output.

5. Civitai Publishing Module: A dedicated group streamlines publishing on Civitai by automatically incorporating essential generation details, such as:

   • Model used

   • CFG (Guidance) value

   • Sampler selected

   • Prompts (positive and negative)

   • Seed

   • Resolution

Model Placement Instructions:

• Checkpoint: Place in path to comfyui/models/checkpoints

• LoRA Models: Place in path to comfyui/models/loras

• BBox Models: Place in path to comfyui/models/ultralytics/bbox

• SAM Models: Place in path to comfyui/models/sams

• Upscaler Models: Place in path to comfyui/models/upscale_models

Customizable Configurations

This workflow offers adjustable settings, including model selection, steps, CFG, sampler, and scheduler to meet specific needs for each image generation project.

Custom Nodes

• https://github.com/kijai/ComfyUI-KJNodes

• https://github.com/ltdrdata/ComfyUI-Impact-Pack

• https://github.com/jags111/efficiency-nodes-comfyui

• https://github.com/WASasquatch/was-node-suite-comfyui

• https://github.com/ssitu/ComfyUI_UltimateSDUpscale

• https://github.com/Suzie1/ComfyUI_Comfyroll_CustomNodes

• https://github.com/wallish77/wlsh_nodes

• https://github.com/rgthree/rgthree-comfy

• https://github.com/giriss/comfy-image-saver

• https://github.com/cubiq/ComfyUI_essentials

• https://github.com/pikenrover/ComfyUI_PRNodes

Troubleshooting:

If you experience issues with the LoRA Stacker nodes, right-click on each "LoRA Stacker" node and select "Reload Node." Sometimes, refreshing your browser or even restarting ComfyUI may be necessary to resolve these issues.