How to beginner · 3 min read

How to run Stable Diffusion on RunPod

Q: How to run Stable Diffusion on RunPod

Run Stable Diffusion on RunPod by creating a serverless endpoint or pod with a GPU, then invoking it via the runpod Python SDK or REST API. Use the runpod.Endpoint class to send prompts and receive generated images programmatically.

Quick answer

Run Stable Diffusion on RunPod by creating a serverless endpoint or pod with a GPU, then invoking it via the runpod Python SDK or REST API. Use the runpod.Endpoint class to send prompts and receive generated images programmatically.

PREREQUISITES

Python 3.8+
RunPod API key
pip install runpod

Setup

Install the runpod Python package and set your API key as an environment variable. You also need to create a Stable Diffusion pod or endpoint on the RunPod platform with GPU support.

bash

pip install runpod

export RUNPOD_API_KEY=os.environ["RUNPOD_API_KEY"]

output

$ pip install runpod
Collecting runpod
  Downloading runpod-1.0.0-py3-none-any.whl (10 kB)
Installing collected packages: runpod
Successfully installed runpod-1.0.0

$ export RUNPOD_API_KEY="your_runpod_api_key_here"

Step by step

Use the runpod Python SDK to invoke your Stable Diffusion endpoint. Replace YOUR_ENDPOINT_ID with your actual RunPod endpoint ID. The example sends a prompt and receives a generated image URL.

python

import os
import runpod

runpod.api_key = os.environ["RUNPOD_API_KEY"]

# Replace with your actual RunPod Stable Diffusion endpoint ID
endpoint = runpod.Endpoint("YOUR_ENDPOINT_ID")

input_data = {
    "prompt": "A fantasy landscape with mountains and a river, digital art",
    "num_inference_steps": 50,
    "guidance_scale": 7.5
}

result = endpoint.run_sync({"input": input_data})

print("Generated image URL:", result["output"])

output

Generated image URL: https://runpod-public.s3.amazonaws.com/generated-images/abc123.png

Common variations

You can use asynchronous calls with endpoint.run_async() for non-blocking requests. Adjust parameters like num_inference_steps and guidance_scale to control image quality and style. You can also switch to different Stable Diffusion versions by selecting the appropriate pod on RunPod.

python

import asyncio

async def generate_image():
    runpod.api_key = os.environ["RUNPOD_API_KEY"]
    endpoint = runpod.Endpoint("YOUR_ENDPOINT_ID")
    input_data = {
        "prompt": "A futuristic cityscape at sunset",
        "num_inference_steps": 30,
        "guidance_scale": 8.0
    }
    result = await endpoint.run_async({"input": input_data})
    print("Generated image URL:", result["output"])

asyncio.run(generate_image())

output

Generated image URL: https://runpod-public.s3.amazonaws.com/generated-images/def456.png

Troubleshooting

If you get authentication errors, verify your RUNPOD_API_KEY environment variable is set correctly.
If the endpoint ID is invalid, double-check your RunPod dashboard for the correct Stable Diffusion pod or endpoint ID.
For slow responses, ensure your pod has GPU resources allocated and is running.

✅

Key Takeaways

Use the official runpod Python SDK with your API key to interact with Stable Diffusion pods.
Set up a GPU-enabled Stable Diffusion pod on RunPod before invoking it programmatically.
Adjust inference parameters like num_inference_steps and guidance_scale to customize image output.
Use asynchronous calls for non-blocking image generation in Python.
Check environment variables and endpoint IDs carefully to avoid common errors.

Verified 2026-04 · stable-diffusion-v1, stable-diffusion-v2

Verify ↗