Stability AI doubles down on AWS

Microsoft may have long had OpenAI as its trusty partner (after its sizable investment), but AWS today announced that Stability AI, one of the hottest new upstarts in the generative AI space and the company behind Stable Diffusion, is doubling down on its cloud, making it its “preferred cloud provider to build and scale its AI models for image, language, audio, video, and 3D content generation.”

In addition, Stability AI will also work with AWS to make its open source tools and model available to more students, researchers, startups and enterprises (which sounds quite a bit like what Microsoft and OpenAI said when they announced their partnership).

Stability AI, which recently announced a $101 million funding round at a valuation of over $1 billion, was already using thousands of Nvidia GPUs in the AWS cloud to train its models. Now, the two companies are formalizing this relationship, with Stability AI planning to use AWS’ SageMaker ML platform, on top of its lower-level infrastructure services with GPUs and AWS’ own Trainium chips.

Typically, these deals also come with preferred pricing and other perks, though neither AWS nor Stability mention those in today’s announcement.

“At Stability AI, our mission is to build the foundation to activate humanity’s potential through AI,” said Emad Mostaque, founder and CEO of Stability AI. “AWS has played an integral role in scaling our open-source foundation models across modalities. We are delighted to run these models on Amazon SageMaker to enable tens of thousands of developers and millions of users to leverage the power of AI with a robust set of tools. We look forward to seeing the amazing things that developers build and customers design and implement using collective intelligence and augmented technology.”

Models like Stable Diffusion, which recently hit version 2.0, are not without controversy, be that for their ability to generate adult content, something Stable Diffusion 2.0 automatically filters out, or because these models are often trained on images from artists who have not explicitly opted in to their works being used to train these models. There can be no doubt, though, that Stability AI is taking over quite a bit of mindshare from OpenAI these days, which has long taken a more cautious approach to how it exposes its models. Stability AI’s open source approach, at least for the time being, seems to be winning in bringing on more developers and — for better or worse — driving innovation in this space.