Skip to content

DiffusionKit

Run diffusion models locally on Apple silicon

Visit official site

What DiffusionKit is

The project is described as an open source toolkit for local diffusion model inference on Apple silicon, converting PyTorch models into Core ML format.

What you can do with it

  • Convert a PyTorch model to Core ML
  • Generate an image with MLX and tune parameters
  • Transform an existing image into another style

Setup

  • Install the Python environment and dependencies
  • Accept the Stable Diffusion 3 terms on Hugging Face Hub
  • Load a local checkpoint when a custom model is needed

Pros & cons

✓ What we like

  • Python CLI and Swift code are both available
  • Local checkpoints let users load custom models

! What to watch out for

  • Users install the Python environment and dependencies
  • PyTorch models must be converted to Core ML

FAQ

Which models are supported?

The material names Stable Diffusion 3 and FLUX.

How is generation run?

Through a Python CLI or Swift code, using MLX.

Last reviewed: 2026-11-15

More AI image generation tools

View all →

How we review