Generate Images from Code

Enable ComfyUI server mode

python main.py --listen --port 8188

Queue a prompt from Python

import requests, json

with open("workflow.json") as f:
    prompt = json.load(f)

# Update the positive prompt node (node IDs depend on your workflow)
prompt["3"]["inputs"]["text"] = "a cyberpunk cat wearing goggles"

r = requests.post("http://127.0.0.1:8188/prompt", json={"prompt": prompt})
print(r.json())

Wait for completion

ComfyUI is asynchronous. Poll the history endpoint:

import time

prompt_id = r.json()["prompt_id"]
while True:
    history = requests.get(f"http://127.0.0.1:8188/history/{prompt_id}").json()
    if history:
        break
    time.sleep(1)

Batch generation

Loop over a list of prompts and queue them:

prompts = ["a red car", "a blue car", "a green car"]
for p in prompts:
    prompt["3"]["inputs"]["text"] = p
    requests.post("http://127.0.0.1:8188/prompt", json={"prompt": prompt})

Use cases

Generate app assets and icons
Create synthetic training datasets
Produce thumbnails and marketing images
Build internal image-generation tools

Tip

Save your ComfyUI workflow JSON with the API format option so node IDs are stable.