What is Easy Diffusion?
Easy Diffusion is an open source app that installs Stable Diffusion on your computer in one step and gives you a simple browser interface for making images. It works on Windows, Linux and Mac, with Nvidia or AMD graphics, Apple silicon, or even just the CPU.
You can create images from text or from other images, inpaint parts of a picture, use ControlNet, embeddings and LoRAs, and fix faces or upscale results with built-in GFPGAN and RealESRGAN. A task queue and live preview make it easy to line up several jobs. The newer engine adds support for models such as Flux.
Main Features
- One-click installer for Windows, Linux and Mac
- Text to image, image to image and inpainting
- ControlNet, LoRA and embedding support
- Face fixing and upscaling built in
- Task queue with live preview
- Runs on Nvidia, AMD, Apple silicon or CPU
How to Get Started with Easy Diffusion
- Download the installer from easydiffusion.github.io.
- On Windows run the installer, or on Linux and Mac unzip it and run start.sh.
- Open the browser interface that launches.
- Type a prompt, choose a model and click Make Image.
Easy Diffusion Pricing
Easy Diffusion is free and open source. It needs at least 8 GB of RAM and about 25 GB of free disk space.
Who Should Use It?
- Beginners who want local AI art without command lines
- Mac users with Apple silicon
- People with AMD graphics cards
- Artists who want private, offline image generation
- Hobbyists trying many models
Easy Diffusion Alternatives
- DiffusionBee AI: Free Stable Diffusion app for Mac.
- Civitai: Community hub for Stable Diffusion models.
- Stable Diffusion XL: Online SDXL image generator.
Frequently Asked Questions
Is Easy Diffusion free?
Yes. Easy Diffusion is free and open source, and it runs on your own computer.
Does Easy Diffusion need an Nvidia GPU?
No. It supports Nvidia and AMD cards, Apple silicon Macs and CPU-only mode, though a GPU is much faster.
Can it upscale and fix faces?
Yes. RealESRGAN upscaling and GFPGAN face correction are built in.