Amid much anticipation, the veil has finally been lifted on the early preview of Stable Diffusion 3. This groundbreaking text-to-image model promises significant improvements in areas such as multi-subject prompts, image quality, and spelling precision, marking a substantial leap forward in artificial intelligence technology.

You can sign up to join the waitlist here.

Although the model is not yet widely available, we are delighted to announce the opening of the early access waitlist today. This phase is crucial for gathering invaluable insights to further elevate its performance and safety, laying a solid foundation for its official release. Join the waitlist now for a firsthand experience of the future of text-to-image technology.

The Stable Diffusion 3 suite encompasses models ranging from 800M to 8B parameters, demonstrating our commitment to accessibility and quality. This diverse range empowers users to choose scalability options tailored to their creative needs. Leveraging a fusion of diffusion transformer architecture and flow matching technology, Stable Diffusion 3 delivers unparalleled performance. Stay tuned for Stability.ai forthcoming detailed technical report to gain deeper insights into its operational principles.

At the core of mission lies the principles of safety and responsible AI application. Rigorous measures have been implemented at every stage of development to prevent potential misuse.

From training and testing to deployment, safety remains paramount. Ahead of this early preview, Stability.ai have introduced a variety of safeguards to ensure the responsible utilization of Stable Diffusion 3. Through ongoing collaboration with researchers, experts, and our community, Stability.ai pledge to maintain integrity as their approach the public release of the model and continue to drive innovation in artificial intelligence.

For All Stable Diffusion Lovers

Thank you for being a part of the Stable Diffusion community!