Developer API Overview

Programmatic access to ComfyUI Cloud for building applications and automations

The ComfyUI Cloud Developer API provides a clean, SDK-friendly interface for running ComfyUI workflows programmatically. Use it to build applications, batch processing pipelines, or integrate AI image generation into your products.

Why Use the Developer API?

Clean REST Semantics

Designed for developers, not constrained by ComfyUI frontend compatibility

Official SDKs

Type-safe clients for Python, JavaScript, and Go

Real-time Events

WebSocket streaming for live progress updates

Webhooks

Push notifications when jobs complete

Core Concepts

Resources

Resource	Description
Inputs	Files you upload for use in workflows (images, masks, etc.)
Models	Custom models you bring (BYOM - Bring Your Own Model)
Jobs	Workflow executions
Outputs	Generated files from completed jobs
Archives	Bulk ZIP downloads of multiple job outputs
Webhooks	Event notifications to your server

Async Patterns

Many operations are asynchronous:

1. Create - Returns 202 Accepted with resource in pending state
2. Poll - Check status until ready or failed
3. Use - Access the resource once ready

# Example: Upload from URL (async operation)
input = client.inputs.from_url("https://example.com/image.png")
# Status is "downloading"

# Poll for completion
while input.status == "downloading":
    time.sleep(1)
    input = client.inputs.get(input.id)

# Now ready to use
print(input.name)  # Use this in your workflow

Or use WebSockets for real-time updates without polling.

Quick Example

from comfy_cloud import ComfyCloudClient

client = ComfyCloudClient(api_key="comfyui-...")

with client:
    # Run a workflow
    job = client.jobs.create(
        workflow={
            "3": {
                "class_type": "KSampler",
                "inputs": {"seed": 42, "steps": 20, ...}
            },
            # ... rest of workflow
        },
        tags=["my-app", "batch-1"],
    )
    
    # Wait for completion
    while job.status == "pending":
        job = client.jobs.get(job.id)
    
    # Get outputs
    outputs = client.outputs.list(job_id=job.id)
    for output in outputs.outputs:
        print(f"Download: {output.download_url}")

Next Steps

Authentication

Get your API key and authenticate requests

Python SDK

Full Python SDK reference and examples

WebSockets

Real-time event streaming

Webhooks

Push notifications for job events