AI startup Stability AI has launched a series of new models for image generation, dubbed Stable Diffusion 3.5.
Introducing Stable Diffusion 3.5, our most powerful models yet.
This open release includes multiple variants that are highly customizable for their size, run on consumer hardware, and are free for both commercial and non-commercial use under the permissive Stability AI Community… pic.twitter.com/KlyE8OjrxN
— Stability AI (@StabilityAI) October 22, 2024
The following versions have been introduced:
- Stable Diffusion 3.5 Large — the flagship model, boasting 8 billion parameters and “superior quality.” It is suitable for professional use with a resolution of 1 megapixel.
- Stable Diffusion 3.5 Large Turbo — an enhanced version of Large, capable of producing images at high speed, albeit with some loss of quality.
- Stable Diffusion 3.5 Medium — featuring 2.5 billion parameters, improved MMDiT-X architecture, and training methods. Designed to work “out of the box” on consumer hardware, it can generate images with resolutions ranging from 0.25 to 2 megapixels. The model will be available on October 29.
The new version of the image generator is expected to produce more diverse results without extensive prompts, according to the company’s announcement. It also offers easy customization to meet specific creative needs, efficient performance, and a wide range of styles and aesthetics.
“Moreover, our analysis shows that Stable Diffusion 3.5 Large leads the market in efficiency and competes with much larger models in image quality,” the startup’s announcement stated.
Back in February, Stability AI introduced the third generation image generator.