Stability AI, the startup behind Stable Diffusion, the tool that uses generative AI to create images from text prompts, revealed Stable Diffusion 3, the next-generation model, on Thursday. Stability AI claims that the new model, which is not yet widely deployed, improves image quality, works better with prompts containing multiple objects, and can more accurately include text as part of the generated image, something that previous models Stable Diffusion weren’t great at.

Stability AI CEO Emad Mosque posted some examples of this on X.

The announcement comes days after Stability AI’s biggest rival, OpenAI, unveiled Sora, a brand new AI model capable of generating near-realistic, high-definition videos from simple text prompts. Sora, which is also not yet available to the general public, has raised concerns about its potential to create realistic-looking fake footage. OpenAI said it is working with experts on disinformation and hateful content to test the tool before making it widely available. Stability AI said it does the same. “[We] have taken and continue to take reasonable steps to prevent abuse of Stable Diffusion 3 by bad actors,” the company wrote in a blog post on its website. “By continuously collaborating with researchers, experts and our community, we expect to innovate with integrity as we move closer to the model’s public release.”

It’s not clear when Stable Diffusion 3 will be released to the public, but until then anyone can join waiting list.

https://www.engadget.com/stable-diffusion-3-is-a-new-ai-image-generator-that-wont-mess-up-text-in-pictures-its-makers-claim-233751335.html?src=rss