Stable Diffusion has introduced the new AI model Stable Diffusion 3 Medium

Stable Diffusion launched the new SD 3 Medium model, which boasts impressive features and innovations. It represents a real breakthrough in the evolution of generative AI.

The CD3 model includes 2 billion parameters, which ensures the generation of high-quality realistic images and perfect character recognition.
The updated model is ideal for running on standard GPUs, and versions optimized for Tensor RT provide better performance.

What other features STD 3 has:
- Elimination of facial inaccuracies and hand distortions
- Better recognition of prompts and styles
- With the help of the Diffusion Transformer architecture, the text in the pictures becomes clear without artifacting and spelling errors.
- With its capacity to absorb nuances from small data sets, it is perfect for tuning purposes.

Collaboration with AMD has optimized the output for SD3 Medium for a variety of AMD devices, including AMD's latest APUs, consumer GPUs, and MI-300X enterprise GPUs.

Stability AI plans to improve Stable Diffusion 3 Medium with feedback, as well as expand its functionality and performance.