Stable Diffusion
by Stability AI
FreemiumOpen-source text-to-image model offering high-quality image generation with extensive customization, LoRA support, and local deployment.
Overview
Stable Diffusion is Stability AI's open-source text-to-image generation model, the most widely used AI image model in the world. Version 3.5 (January 2026) features enhanced prompt adherence, diverse outputs, improved text rendering, better composition, and ControlNet support for precise image control.
Being open-source, Stable Diffusion can be downloaded and run locally on consumer GPUs, fine-tuned with custom LoRA models, and integrated into any pipeline. It supports resolutions up to 1 megapixel and generates everything from photorealism to artistic renders. The ecosystem includes thousands of community models, extensions, and GUIs (ComfyUI, Automatic1111, Forge).
Available as open-weight downloads, through the DreamStudio web app, or via the Stability API for developers.
Pricing
- Download open-weight models from Hugging Face
- Run locally on your own GPU
- Community License free for individuals and businesses under $1M annual revenue