Stable Diffusion is an open-source AI image generation model originally developed by Stability AI in collaboration with academic research groups, first released in 2022. It was among the first high-quality text-to-image models to be made publicly available with open weights, meaning anyone could download, run, and modify the model, which sparked a massive ecosystem of community development, fine-tuned variants, and derivative tools.

The model uses a latent diffusion architecture that performs the denoising process in a compressed latent space rather than directly on pixels, making it significantly more computationally efficient than earlier diffusion approaches and enabling generation on consumer hardware rather than requiring large cloud compute resources. This accessibility, combined with its open-source nature, led to an explosion of community-created fine-tuned models, LoRAs, extensions, and interfaces built on top of the base model. Platforms like Automatic1111, ComfyUI, and many others emerged as user interfaces for running Stable Diffusion locally, and the model became the foundation for a large portion of the open-source AI image generation ecosystem. Multiple major versions have been released since the original, each bringing improvements in quality, prompt adherence, and capability.

Stable Diffusion's open-source release was a defining moment in the democratization of AI image generation, establishing that high-quality generative models could be run outside of large company infrastructure and enabling a global community of developers, artists, and researchers to build on, study, and extend the technology. Many commercial AI generation tools are built on or directly inspired by Stable Diffusion's architecture and open development model.