The world of artificial intelligence (AI) is constantly evolving, and the recent release of the new image model from Stable Diffusion is a testament to that. The innovators at Stability AI have made headlines again, this time with the release of SDXL 0.9, a groundbreaking development in their suite of text-to-image models. This model offers a significant improvement in image composition and detail over its predecessor.

The Power of SDXL 0.9

Don’t let the technical name fool you. SDXL 0.9 is more than a collection of algorithms and numbers; it’s a revolution in the realm of generative AI imagery. And despite its advanced features, you don’t need a supercomputer to run it. You can run SDXL 0.9 on a modern consumer GPU, provided you have a Windows 10 or 11, or Linux operating system, with 16GB RAM, and an Nvidia GeForce RTX 20 graphics card (or higher standard) with at least 8GB of VRAM. Linux users can also use a compatible AMD card with 16GB VRAM.

A Spectrum of Applications

The ability of SDXL 0.9 to generate hyper-realistic creations extends its applications beyond traditional bounds. It can be used for films, television, music, and instructional videos, as well as design and industrial uses. This places SDXL at the forefront of real-world applications for AI imagery.

Going Beyond Basic Text Prompting

One of the notable features of the SDXL series is its range of functionalities. It isn’t limited to basic text prompting. The model offers functionalities such as image-to-image prompting (inputting one image to get variations of that image), inpainting (reconstructing missing parts of an image), and outpainting (creating a seamless extension of an existing image).

What Makes SDXL 0.9 Stand Out?

The key driver of this advancement in composition for SDXL 0.9 is its significant increase in parameter count (the sum of all the weights and biases in the neural network that the model is trained on) over the beta version. SDXL 0.9 boasts one of the largest parameter counts of any open-source image model. It has a 3.5B parameter base model and a 6.6B parameter model ensemble pipeline. This means that the final output is created by running on two models and aggregating the results. The second stage model of the pipeline adds finer details to the output of the first stage.

Additionally, SDXL 0.9 runs on two CLIP models, including one of the largest OpenCLIP models trained to date. This enhances its processing power and ability to create realistic imagery with greater depth and a higher resolution of 1024×1024.

The Community Response

Since the beta launch of SDXL in April, the response from the user community has been astounding. The model has generated more than 700,000 images, averaging more than 20,000 per day. More than 54,000 images have been entered into community ‘Showdowns’ with 3,521 SDXL images nominated as winners.

Availability and Future Plans

SDXL 0.9 is now available on the Clipdrop by Stability AI platform. API and DreamStudio customers, as well as users of other leading image generating tools like NightCafe, will be able to access the model. During a limited period, SDXL 0.9 will be providedfor research purposes only to collect feedback and refine the model before its general open release. The code to run it will be publicly available on Github.

The future looks bright for SDXL. The full open release of SDXL 1.0 is targeted for mid-July, marking another milestone in AI image generation. Keep in mind, though, that SDXL0.9 is released under a non-commercial, research-only license and is subject to its terms of use.

Conclusion

The new image model from Stable Diffusion, SDXL 0.9, represents a significant step forward in AI image generation. It offers vastly improved image composition and detail, a range of creative applications, and ease of use on modern consumer GPUs. With the capacity to generate hyper-realistic creations and a host of advanced functionalities, SDXL 0.9 is truly at the forefront of AI imagery.

FAQs

1. What is the new image model from Stable Diffusion?
   • The new image model from Stable Diffusion is SDXL 0.9, developed by Stability AI. It is a significant advancement in the suite of text-to-image models, offering improved image composition and detail.
2. What are the system requirements for running SDXL 0.9?
   • SDXL 0.9 can be run on a modern consumer GPU, needing only a Windows 10 or 11, or Linux operating system, with 16GB RAM, and an Nvidia GeForce RTX 20 graphics card (or higher standard) equipped with a minimum of 8GB of VRAM. Linux users can also use a compatible AMD card with 16GB VRAM.
3. What are the applications of SDXL 0.9?
   • SDXL 0.9 can be used to generate hyper-realistic creations for films, television, music, instructional videos, design, and industrial uses. It also offers functionalities that extend beyond basic text prompting, like image-to-image prompting, inpainting, and outpainting.
4. When will the full open release of SDXL 1.0 be available?
   • The full open release of SDXL 1.0 is targeted for mid-July.