MULTI_REGION_WORKFLOW
Unlock Your Creative Potential with Two Unique Workflows for ComfyUI
Imagine having complete control over the composition of your images, creating results that perfectly reflect your creative vision. With these two workflows designed for Regional Prompting, now you can!
✅ Workflow 1: Leverage Flux Redux to precisely transfer the style and composition of reference images. The result? Faithful, personalized creations that are ideal for achieving an accurate aesthetic.
✅ Workflow 2: Powered by Florence 2, a cutting-edge VLM model, this workflow interprets images and translates them into complex textual prompts, ensuring both coherence and flexibility for intricate compositions.
🎨 Expand, customize, create: These workflows are the perfect starting point for anyone seeking advanced control and a solid foundation for experimentation.
💡 Don’t let complexity hold you back! Take the first step toward extraordinary compositions.
Instructions:
To use the workflows, simply open or drag the .json
file into ComfyUI. You’ll receive a notification about any missing nodes. Navigate to the Manager and select Install missing custom nodes to automatically install everything you need to make the workflows function properly.
You’ll also need the following models:
- Flux Dev
- Flux Redux
- Florence 2
- Clipvision
- ClipText Encoder
- Flux VAE
You can search for and download these models directly from the Manager.
Important Note on Nodes:
Pay close attention to the following nodes:
ControlAltAI Nodes - GitHub Repository
Make sure to read the instructions provided on the page. It’s essential to install Xformers for these nodes to function correctly. You can find detailed installation instructions here:
Xformers Installation Instructions
MULTI_REGION_BY_PROMPT.json, MULTI_REGION_BY_IMAGES.json