# MochiDiffusion/MochiDiffusion

Run Stable Diffusion on Mac natively

Repository: https://github.com/MochiDiffusion/MochiDiffusion
Canonical: https://ross.abutalabs.com/products/mochidiffusion
Language: Swift
License: GPL-3.0
License Family: copyleft
Topics: apple, apple-silicon, macos, stable-diffusion, coreml, ane, neural-engine, swift, swiftui
Last push: 2026-08-23T00:12:28+00:00

## Health v2 (maintenance only)
Score: 78/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 40, longevity 96
- inputs: {"age_days": 1357, "days_push": 11, "days_rel": 186, "gap_med": 486, "n_releases_24m": 2}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 7926, forks 364 (observed 2026-08-28T04:10:09.864808+00:00)

## What it is
A native macOS app built with SwiftUI for running Stable Diffusion and FLUX.2 Klein image generation locally on Apple Silicon Macs. It uses Apple's Core ML Stable Diffusion implementation to leverage the Neural Engine for fast, memory-efficient, fully offline image generation.

## Use cases
- generate images from text prompts on a Mac
- run Stable Diffusion locally and offline on Apple Silicon
- convert text to images using the Neural Engine
- do image-to-image generation with ControlNet
- generate AI images without a GPU cloud service
- run FLUX models natively on macOS

## When to choose
- you have an Apple Silicon Mac and want fast, offline Stable Diffusion
- you want low memory usage (~150MB) via the Neural Engine
- you prefer a native macOS GUI over command-line tools
- you want ControlNet and image-to-image support with EXIF prompt metadata

## When to avoid
- you are on Linux or Windows
- you need CUDA/NVIDIA GPU acceleration
- you want to integrate diffusion into a Python pipeline or server backend
- you need models not convertible to Core ML format

## Facets
- artifact type: application
- maturity: active
- function: image-processing, machine-learning, gui
- domain: image-processing, machine-learning, apple-ecosystem
- platform: -
- tags: stable-diffusion, coreml, apple-silicon, neural-engine, text-to-image, image-generation, swiftui, flux, controlnet, offline, macos, swift

## Member repositories
- MochiDiffusion/MochiDiffusion (main) score 78

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:09.864808+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:33:15.420567+00:00, confidence not recorded.
  - readme: https://github.com/MochiDiffusion/MochiDiffusion (fetched 2026-08-28T04:10:09.864808+00:00, sha a506cabd3694)
- Data as of 2026-08-30T08:39:29.467469+00:00.