OpenAI announced that its frontier models, including its Codex code-generation model, are now generally available on Amazon Web Services, according to a post on OpenAI's official blog. The integration allows AWS customers to access OpenAI's models directly through their existing cloud accounts rather than setting up separate contracts.
According to OpenAI, the change lets businesses use their existing AWS billing, identity and access management, and security controls when deploying OpenAI's models. This is intended to reduce the procurement and integration work previously required to move from testing a model to running it in production.
Developers can incorporate Codex and other OpenAI models into applications running on AWS using the same tools and workflows they already use for other AWS services, per the announcement. OpenAI frames this as easing adoption for regulated industries such as finance and healthcare, where staying within an approved cloud environment for data governance is often a requirement.
The move expands OpenAI's distribution beyond its existing relationship with Microsoft Azure, giving AWS's enterprise customer base direct access to its models without a separate integration layer. OpenAI's announcement does not specify pricing details beyond noting it will run through AWS billing, nor which specific frontier models are included beyond Codex.