Stability AI, a company that develops generative AI models for text and image generation has launched a new version of its image generation engine, Stable Diffusion XL 1.0. The engine is based on a large-scale neural network model that can produce high-quality and AI-generated images and art. It is currently available for testing on the Discord bot before coming to DreamStudio.
Stable Diffusion is a generative AI model that uses a diffusion process to gradually transform random information into coherent images. Programs like Stable Diffusion can generate images of various categories, such as animals, landscapes, portraits, and abstract art, as well as images that combine multiple categories or concepts.
Stable Diffusion XL (SDXL) enables you to generate expressive images with shorter prompts and insert words inside images. The model is a remarkable improvement in image generation abilities, offering enhanced image composition and face generation that result in stunning visuals and realistic aesthetics.
It has been in Beta testing until now. Stability AI announced on its Stability Foundation Discord channel and on Twitter that Stable Diffusion XL is now graduating to version 1.0. Like the standard Stable Diffusion, SDXL runs on the Stable Studio from Stability AI.
Ready, set, 🚀!
Our SDXL Bot is now 💫LIVE💫 with final candidates for SDXL 1.0!
Hop onto our bot channels to test, vote and help us pick the 🌟 for the big release! pic.twitter.com/IM6KPJWU3Q
— Stable Foundation (@StableDiffusion) July 10, 2023
New Abilities and Features of Stable Diffusion XL
- Faster and more stable generation: The engine can understands user text prompts more quickly and reliably, with lower memory consumption and better error handling. The engine can also handle longer and more complex inputs and outputs without compromising quality or diversity.
- More control and customization: Users can adjust the tone, style, length, and format of their prompts according to their preferences and needs. Users can also specify keywords, phrases, or sentences to guide the generation process or to include in the output.
- More feedback and guidance: The engine provides users with more information and feedback on the generated content, such as the progress, confidence, and relevance scores.
- More integration and compatibility: The engine can work with other platforms and tools that users may use for content creation or data analysis, such as Discord, Google Docs, WordPress, etc. The engine can also import and export data from various sources and formats, such as text files, CSV files, JSON files, etc.
Now Available on Discord and Coming to Stable Studio
Stability is currently allowing users on the Foundation Discord to test the SDXL bot. It will also come to the DreamStudio. If you are unfamiliar with DreamStudio, it was the pioneer consumer app that demonstrated the power of Stable Diffusion. It debuted in August 2022 and has since drawn millions of users who have made and shared millions of images using the app.
DreamStudio also boasts a gallery of selected images, a community forum, and a premium subscription option that provided higher resolution and faster generation. In May, Stability AI chose to launch an open-source version of DreamStudio, called StableStudio, in order to encourage cooperative and transparent development of the app.