Stability AI is continually developing their generative AI models, despite the rising competition and moral issues that come along with it. Joe Penna, the company’s head of applied ML, mentioned to TechCrunch that Stable Diffusion XL 1.0 – comprised of 3.5 billion parameters – can generate 1-megapixel resolution images “in seconds” and in a range of aspect ratios. Parameters are components of a model taught using data samples and fundamentally define its skill when it comes to producing images.