Startup Stability AI has given users access to its latest neural network, Stable Diffusion 3, which creates images based on given queries. Currently, Stable Diffusion 3 can only be used through the company's API.
Users can submit queries to generate images, but do not yet have the ability to run the model on their hardware or refine it using their own data. Stability AI has not announced a timeline for a full-fledged release of the model.
Company representatives state that in terms of query accuracy Stable Diffusion 3 either outperforms or is on par with such systems as DALL-E 3 and Midjourney V6. At the same time, the quality of image generation has improved markedly. Stable Diffusion 3 is based on the Multimodal Diffusion Transformer architecture and features separate toolkits for image and text processing, which has significantly improved the understanding and generation of text queries.