Stable Diffusion, the prompt-based AI image generator, is getting an upgrade that promises to eliminate the notorious “spaghetti hands” problem. Stability AI has announced the upcoming release of SDXL 0.9, an improved version of Stable Diffusion XL, which is expected to produce highly realistic images and greatly enhance the quality of hand renderings.
Stability AI, in a now-deleted blog post later reported by Bloomberg, revealed that SDXL 0.9 offers significant advancements in image and composition detail compared to its predecessor. The blog post showcased sample images generated by Stable Diffusion XL beta and SDXL 0.9 using identical prompts. The results clearly demonstrated the newer model’s superior ability to capture intricate details, such as aliens, a wolf, and a person holding a coffee cup, with noticeably improved and more convincing hand representations. Previously, the quality of hands in AI-generated art had often been a giveaway, but the release of Midjourney v5 on Discord in March had partly addressed this limitation.
Stability AI enthusiastically described the new model as a game-changer for creative applications of generative AI imagery. The company stated that SDXL 0.9 would empower artists, filmmakers, TV producers, musicians, instructional video creators, designers, and industrial users by enabling them to generate hyper-realistic visuals. Stability AI positioned SDXL as a leading solution for real-world implementations of AI imagery.
To achieve these enhanced results, Stability AI attributed SDXL 0.9’s success to a significant increase in the model’s parameter count, encompassing all the weights and biases within the neural network it’s trained on. However, running the new version locally on a PC will require a minimum of 16GB of RAM and an 8GB VRAM GeForce RTX 20 series graphics card or higher. Compatibility is provided for Windows 11 / 10 and Linux operating systems.
The now-deleted blog post also mentioned upcoming integration plans for SDXL 0.9. Stability AI intends to make the model available on its Clipdrop web tool and incorporate it into the company’s DreamStudio app. Additionally, an open-source version, SDXL 1.0, is scheduled for release in mid-July, further expanding accessibility and possibilities for users.
Frequently Asked Questions (FAQs) about AI image generator
What is Stable Diffusion XL 0.9?
Stable Diffusion XL 0.9 is the next version of a prompt-based AI image generator called Stable Diffusion. It is an improved model that aims to produce more photorealistic images and overcome the issue of “spaghetti hands” commonly found in AI-generated art.
How does SDXL 0.9 compare to its predecessor?
According to Stability AI, the company behind SDXL, the new model offers significantly improved image and composition detail compared to its predecessor. It has undergone advancements in generating finer details and more convincing hand renderings, making the generated images more realistic.
What are the benefits of SDXL 0.9?
SDXL 0.9 brings a leap in creative use cases for generative AI imagery. It opens up possibilities for various industries, including films, television, music, instructional videos, design, and industrial applications. The model allows for the creation of hyper-realistic visuals, enhancing the quality and realism of AI-generated content.
What are the system requirements for running SDXL 0.9?
To run SDXL 0.9 locally on a PC, a minimum of 16GB of RAM and a GeForce RTX 20 (or higher) graphics card with 8GB of VRAM is required. The model supports Windows 11/10 and Linux operating systems.
When will SDXL 0.9 be available?
According to the information provided, SDXL 0.9 is expected to be available on Stability AI’s Clipdrop web tool soon. It will also be added to the company’s DreamStudio app. Additionally, an open-source version, SDXL 1.0, is scheduled for release in mid-July, which will further expand accessibility and possibilities for users.