OpenAI models, Codex, and Managed Agents come to AWS

Source: OpenAI Date: 2026-04-28 URL: https://openai.com/index/openai-on-aws

Summary

OpenAI’s April 2026 announcement of availability on AWS — bringing GPT-5.x models, Codex, and Managed Agents (OpenAI’s hosted agent infrastructure) to Amazon Bedrock and Amazon SageMaker. The AWS deployment made OpenAI’s full product stack accessible to enterprises that had standardized on AWS infrastructure, without requiring cross-cloud data movement to Azure. This completed the multi-cloud distribution strategy: Azure (existing), AWS (new), and Stargate/Oracle (OpenAI’s own infrastructure).

Implications

AWS as the second major cloud channel. OpenAI’s availability on Bedrock put it in the same position as Anthropic’s Claude — available through the multi-model marketplace that enterprise AWS customers used to evaluate and deploy AI models. The “Managed Agents” inclusion was particularly significant: agentic infrastructure on AWS meant OpenAI’s agent platform could be deployed within AWS VPCs and integrated with AWS services (S3, Lambda, RDS) without data leaving the enterprise’s AWS environment.

Thread: Cloud distribution and multi-cloud strategy. The April 2026 AWS launch was the direct follow-on to the February 2026 Amazon strategic partnership announcement — the partnership was announced, and two months later the actual product availability launched. This is the commercial execution following the strategic agreement.

Watch: Whether OpenAI’s AWS availability cannibalized Azure-hosted OpenAI usage or primarily addressed net-new customers who hadn’t adopted Azure, and how the pricing compared to direct API access through OpenAI.com.