Stability AI Launches Advanced Stable Diffusion XL 1.0 on Amazon Bedrock
Stability AI releases its most advanced text-to-image model, Stable Diffusion XL 1.0, making it available through Amazon Bedrock and as an open-access download. The upgraded model produces vibrant, high-resolution images in seconds with minimal training data.
Stability AI unveils Stable Diffusion XL (SDXL) 1.0 as the most advanced model in its flagship text-to-image suite. The company announces that this powerful new tool is featured on Amazon Bedrock, a fully managed service from AWS that provides access to top foundation models via an API. This launch takes place at the annual AWS Summit New York and highlights a strong partnership between Stability AI and Amazon.
The new model generates highly vibrant and accurate colors with improved contrast, lighting, and shadows compared to its predecessor. It streamlines the imaging process to deliver full 1-megapixel resolution images in just seconds across multiple aspect ratios. These enhancements build upon the data gathered from the research-only release of SDXL 0.9 in June, allowing developers to fine-tune the system for significantly better results.
Developers access SDXL 1.0 immediately through the Stability AI API, GitHub, and consumer applications like Clipdrop and DreamStudio. The model is also available on Amazon SageMaker JumpStart to help machine learning users get started quickly. Additionally, a new fine-tuning beta feature allows users to specialize the image generation for specific people or products using as few as five images.