Stable Diffusion 3 is Stability AI's most capable open image generation model with improved text rendering multi-subject composition and photorealistic quality. It uses a diffusion transformer architecture for superior prompt adherence. Available as open weights for local deployment and via Stability AI API.