Stable Diffusion is a family of open-source diffusion models first released by Stability AI in 2022. Unlike closed systems (DALL-E, Midjourney), the model weights are public and can be downloaded, fine-tuned, and run on a single consumer GPU. That accessibility is why most independent AI interior design tools — including Tigmi — are built on Stable Diffusion derivatives rather than calling out to a paid API.
The practical version most consumer tools use today is Stable Diffusion XL or one of its fine-tuned descendants (Realistic Vision, RealVisXL, Juggernaut). These tuned variants trade some flexibility for sharper photorealism on specific subject matter — interiors, portraits, products — which is what interior design tools actually need.
You rarely interact with Stable Diffusion directly. It sits behind the scenes in image-generation tools, with the user interface layered on top adding controls for style, palette, and composition.