Stable Diffusion 3.5 Enhances Image Diversity and Prompt Adherence
Stability AI's latest models aim to improve image generation quality and accessibility with new features and licensing terms.
Overview
- Stable Diffusion 3.5 introduces three variants: Large, Large Turbo, and Medium, each catering to different user needs and hardware capabilities.
- The models are designed to generate more diverse images, depicting a wider range of skin tones and features without extensive prompting.
- Stable Diffusion 3.5 Large and Large Turbo are available now, while the Medium variant will be released on October 29, 2024.
- Stability AI offers these models under a community license, allowing both commercial and non-commercial use with certain conditions.
- The release addresses past criticisms by improving prompt adherence and reducing visual artifacts, making the models more reliable for creators.