How to beginner · 3 min read

How to run Stable Diffusion on Mac

Q: How to run Stable Diffusion on Mac

Run Stable Diffusion on Mac by installing Python 3.8+, setting up a virtual environment, and installing the diffusers library with torch. Use the StableDiffusionPipeline from diffusers to generate images locally on your Mac.

Quick answer

Run Stable Diffusion on Mac by installing Python 3.8+, setting up a virtual environment, and installing the diffusers library with torch. Use the StableDiffusionPipeline from diffusers to generate images locally on your Mac.

PREREQUISITES

Python 3.8+
pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cpu
pip install diffusers transformers scipy ftfy
Basic terminal and Python knowledge

Setup

Install Python 3.8+ if not already installed. Use pip to install torch (CPU version recommended for Mac), diffusers, and dependencies. Create a virtual environment to isolate packages.

bash

python3 -m venv sd-env
source sd-env/bin/activate
pip install --upgrade pip
pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cpu
pip install diffusers transformers scipy ftfy

Step by step

Use the StableDiffusionPipeline from diffusers to load the model and generate an image. This example uses the runwayml/stable-diffusion-v1-5 model.

python

from diffusers import StableDiffusionPipeline
import torch

# Load the pipeline with CPU device
pipe = StableDiffusionPipeline.from_pretrained(
    "runwayml/stable-diffusion-v1-5",
    torch_dtype=torch.float32
)
pipe = pipe.to("cpu")

# Generate an image from prompt
prompt = "A fantasy landscape, vivid colors"
image = pipe(prompt).images[0]

# Save the image
image.save("output.png")
print("Image saved as output.png")

output

Image saved as output.png

Common variations

Use GPU acceleration on Mac with Apple Silicon by installing torch with MPS support and setting pipe.to("mps").
Run asynchronously with asyncio and diffusers for non-blocking calls.
Try different Stable Diffusion models by changing the model name in from_pretrained.

python

import asyncio
from diffusers import StableDiffusionPipeline
import torch

async def generate_image_async(prompt: str):
    pipe = StableDiffusionPipeline.from_pretrained(
        "runwayml/stable-diffusion-v1-5",
        torch_dtype=torch.float32
    )
    pipe = pipe.to("cpu")
    image = await pipe(prompt).images[0]
    image.save("async_output.png")
    print("Async image saved as async_output.png")

asyncio.run(generate_image_async("A futuristic cityscape at sunset"))

output

Async image saved as async_output.png

Troubleshooting

If you get RuntimeError: CUDA not available on Mac, switch to CPU or MPS device.
For slow performance, ensure you use Apple Silicon optimized torch and MPS device if available.
If diffusers installation fails, upgrade pip and install dependencies manually.

✅

Key Takeaways

Use diffusers and torch CPU or MPS builds to run Stable Diffusion on Mac.
Create a Python virtual environment to manage dependencies cleanly.
Switch device to mps for Apple Silicon Macs to improve performance.
Troubleshoot CUDA errors by using CPU or MPS devices on Mac.
Experiment with different models by changing the from_pretrained argument.

Verified 2026-04 · runwayml/stable-diffusion-v1-5

Verify ↗