Precisely in the hours in which Google paused the generation of Gemini AI images, due to problems related to the creation of historical photos of people, the Stability AI startup presents Stable Diffusion 3.
It is one of the most famous photo-to-text generation tools on the internet, pioneering generative artificial intelligence for the consumer together with the ChatGpt chatbot.
According to an official communication from the developer, the main difference between Diffusion 3 and its predecessors is the possibility for users to also insert texts into the images, in a clear and precise manner, a feature absent elsewhere.
In fact, not infrequently, when using any generative AI tool for photos, the texts in the graphics obtained are confused, with spelling errors and misunderstandings.
At the moment, Stable Diffusion 3 has not been released to the public.
The only way to try it is to sign up to the waiting list and wait for the release of the experimental version.
Other improvements, for Stability AI, include a higher quality of the final images and the opportunity to describe what you want with longer text, also inserting details into the request to ensure that these are present in the artificial images.
The announcement comes just days after Stability AI's biggest rival, OpenAI, unveiled Sora, an AI model that can generate near-realistic, high-definition video from simple text instructions.
Sora is not yet available to the general public.
Reproduction reserved © Copyright ANSA