Python API for Comfy UI with Gradio

This repo contains the code from my YouTube tutorial on building a Python API to connect Gradio and Comfy UI for AI image generation with Stable Diffusion models.

Original project

https://github.com/SharCodin/comfy-gradio-api.git

Overview

This project shows:

• How to connect a Gradio front-end interface to a Comfy UI backend
• Sending workflow data as API requests
• Updating generation parameters dynamically
• Displaying generated images in Gradio
• Adding text and image inputs
• Using a smartphone camera for image inputs

By the end, you'll understand the basics of building a Python API and connecting a user interface with an AI workflow.

Running the Code

To run the code:

1. Clone the repo
2. Install dependencies (pip install requests Pillow gradio numpy)
3. Modify the Comfy UI installation path
4. Open python image_to_image.py and modify the INPUT_DIR and OUTPUT_DIRand' folder path image_to_image.py: resized_image.save(os.path.join(INPUT_DIR, "test_api.jpg"))The "test_api.jpg" is equal to the image of wlopsimple_api.json loading node's "image": "test_api.jpg"
5. Open the localhost link to view in Gradio interface

The main files:

• image_to_image.py - Contains Gradio UI and API logic
• wlopsimple_api.json - Saved Comfy UI workflow

Let me know if you have any other questions!