Stable Diffusion

Stable Diffusion is Stability AI's family of open weight text to image models, and Stable Diffusion 3.5 is the current release. It comes in three variants: Large, an 8.1 billion parameter model built for quality and prompt adherence at 1 megapixel; Large Turbo, a distilled version that generates in four steps; and Medium, a 2.5 billion parameter model designed for consumer hardware at 0.25 to 2 megapixels. The models are used by developers, researchers, small studios and hobbyists who want to run generation on their own hardware, fine tune on their own data or build products without a per image bill. Weights download from Hugging Face and run in ComfyUI and other open tooling; the same models are hosted on Stability's API and on third party services such as Replicate, Fireworks AI and DeepInfra. Stability also sells Stable Image Ultra and Core endpoints, upscalers, editing services, Stable Audio and Stable Fast 3D through the same platform. Licensing follows Stability AI's Community License: free for research, non commercial use and for organizations under $1 million in annual revenue, with outputs owned by the user; above that threshold an Enterprise License is negotiated with Stability. There is no consumer subscription or pricing page. API usage is billed in credits at $0.01 each, with 25 free credits on signup: SD 3.5 Large costs 6.5 credits per image, Large Turbo 4, Medium 3.5, SD 3.5 Flash 2.5, Stable Image Core 3 and Stable Image Ultra 8. Stability AI's home page now leads with licensed media partnerships and enterprise production services; Stable Diffusion remains the open weight core that the rest is built on.