Stability AI, the company behind Stable Diffusion, the industry-leading image generation AI and the foundation of many other text-to-image AI tools, has announced the launch of Stable Cascade, an image generation model that promises to be faster and more powerful than its predecessor.
The new model of Stability AI can generate photos and provide variations of the exact image it has created, or attempt to increase the resolution of an existing photo. Other text-to-image editing functions include “inpainting” and “outpainting”, where the model will only fill in a specific part of the image, as well as “canny edge”, where users can take a new photo using only the edges of an existing image.
The new model is available on GitHub for researchers, but not for commercial use, and it offers more options even when companies like Google and even Apple are releasing their own image generation models.
Stability AI helped popularize the “stable diffusion” method and has also been the subject of several lawsuits alleging that Stable Diffusion was formed using copyrighted data without permission from the rights holders (a trial for one of Getty Images’ lawsuits against Stability AI is scheduled to take place in December in the UK).