Stable Diffusion 3 AI model launched, capable of creating high-quality images from text

Stability AI from a British AI laboratory A new model has been revealed that can create images from new text, Stable Diffusion 3, which can create high-quality images from Prompt text and build on the previous model. Improve work efficiency

The announcement was made after the launch of Sora, a new AI model that can create realistic and detailed videos from Stability AI's main competitor, OpenAI. According to Stability AI, Stable Diffusion 3 is capable of Much better handling multi-object visualizations. Compared to the previous version It allows users to combine more detailed prompts with different elements. lots and get better results

In addition to improvements with complex prompts, The new model also has upgraded overall image quality and spelling accuracy. Stability AI claims these upgrades address consistency and cohesion issues that plagued text-to-image models in the past.

Although not yet publicly available, Stability AI has opened registration for those interested in early access to Stable Diffusion 3. The preview period will allow Stability AI to gather feedback and further refine the model before it is released. A full launch is planned for later this year.

“We believe in safe and responsible AI practices. This means that we have taken and continue to take appropriate steps to prevent the misuse of Stable Diffusion 3 by bad actors. Security begins when we start training our models and continues throughout testing, evaluation, and deployment,” Stability AI said.

“In preparing for this preliminary preview, We have introduced several safeguards. Through continuous collaboration with researchers, experts, and our community. We expect to innovate. Continue to be honest as we move closer to the public release of the model.”

Stable Diffusion 3 is available in a range of sizes, from 800 million parameters on the low end to 8 billion on the high end. Stability AI says that the options aim to strike a balance between creative performance and user access to different computational resources.