Stability AI Launches High-Resolution Stable Diffusion XL 1.0 Model
Stability AI unveils Stable Diffusion XL 1.0, an open-source text-to-image model that generates highly detailed 1-megapixel images with improved color and text generation capabilities.
Stability AI releases Stable Diffusion XL 1.0, calling it the most advanced text-to-image model in its lineup. The open-source model contains 3.5 billion parameters and allows users to generate full 1-megapixel resolution images in seconds across multiple aspect ratios. It is available through GitHub, Stability's API, and consumer apps like ClipDrop and DreamStudio.
The new model brings significant visual improvements, delivering more vibrant and accurate colors along with better contrast, shadows, and lighting than its predecessor. It also excels at generating legible text within images, a common struggle for competing AI models. Furthermore, the system handles complex, multi-part instructions from short prompts instead of requiring lengthy text inputs.
Stable Diffusion XL 1.0 supports advanced features like inpainting, outpainting, and image-to-image prompts for creating detailed variations of existing pictures. The model remains highly customizable and ready for fine-tuning to match specific concepts and styles. However, the open-source release continues to raise familiar moral and ethical questions regarding AI-generated content.