After a long time, I have compiled my personal workflow for generating locally on comfyui. This workflow is primarily designed for flux, but you can modify it according to your needs.

FLUX 1.D — T2IMG / IMG2IMG Workflow for ComfyUI

A modular ComfyUI workflow designed for flexible text-to-image and image-to-image generation using FLUX and LoRA models.
The setup integrates advanced sampler controls, metadata saving, conditioning management, and Florence 2 model support.

🧩 Features

• ✅ Dual text encoding with CLIP + T5

• ✅ Support for LoRA (MODEL + CLIP)

• ✅ Text-to-Image and Image-to-Image workflows combined

• ✅ Advanced control over:

  • Sampler / Scheduler

  • CFG Scale, Steps, Seed

  • Width & Height

• ✅ Metadata saving with full generation parameters

• ✅ Integration with Florence 2 model loader

• ✅ Preconnected nodes for clean extension or modification

⚙️ Requirements

• ComfyUI

• Installed custom nodes:

  • comfyui-kjnodes

  • comfyui-image-saver

  • comfyui-florence2

  • comfy-core

  • Lora manager

---