OpenAI and Amazon Strategic Partnership

OpenAI and Amazon have entered into a multi-year strategic partnership to accelerate AI innovation for enterprises, startups, and consumers. This collaboration includes a $50 billion investment in OpenAI, the development of a Stateful Runtime Environment on Amazon Bedrock, and a massive expansion of OpenAI's compute capacity using AWS Trainium silicon.

$50 Billion Investment and Financial Terms

Amazon will invest $50 billion in OpenAI. This investment is structured as an initial $15 billion payment, followed by an additional $35 billion in the coming months contingent upon certain conditions being met.

Co-Creation of the Stateful Runtime Environment

OpenAI and AWS are co-creating a Stateful Runtime Environment powered by OpenAI models and available via Amazon Bedrock. This environment enables generative AI applications and agents to operate at production scale by providing models with seamless access to compute, memory, and identity.

Unlike traditional stateless interactions, the Stateful Runtime Environment allows developers to:

  • Maintain context across sessions.
  • Remember prior work.
  • Operate across various data sources and software tools.
  • Access dedicated compute for ongoing projects and workflows.

The Stateful Runtime Environment is expected to launch within the next few months.

Exclusive Distribution of OpenAI Frontier on AWS

AWS will serve as the exclusive third-party cloud distribution provider for OpenAI Frontier. Frontier is an enterprise platform designed for organizations to build, deploy, and manage teams of AI agents that operate across business systems with shared context, built-in governance, and enterprise-grade security.

By utilizing AWS as the distribution provider, organizations can integrate AI agents into existing workflows at a global scale without the need to manage the underlying infrastructure.

Infrastructure Expansion and Trainium Compute

OpenAI and AWS are expanding an existing $38 billion multi-year agreement by an additional $100 billion over eight years. As part of this expansion, OpenAI has committed to consume approximately 2 gigawatts of Trainium capacity through AWS infrastructure to support the Stateful Runtime Environment, OpenAI Frontier, and other advanced workloads.

This compute commitment includes the use of both Trainium3 and next-generation Trainium4 chips. Trainium4, which is expected to begin delivery in 2027, is designed to provide performance gains through:

  • Significantly higher FP4 compute performance.
  • Expanded memory bandwidth.
  • Increased high-bandwidth memory capacity.

Custom Model Development for Amazon Applications

OpenAI and Amazon will collaborate to develop customized models specifically for Amazon's customer-facing applications. Amazon developers will be able to tailor OpenAI models for use in AI products and agents that serve customers directly, complementing existing options such as Amazon's Nova family of models.

“OpenAI and Amazon share a belief that AI should show up in ways that are practical and genuinely useful for people. Combining OpenAI’s intelligence with Amazon’s infrastructure and global reach helps us put powerful AI into the hands of businesses and users at real scale.“

— Sam Altman, co-founder and CEO of OpenAI

Sources