OpenAI models and Codex arrive in AWS Bedrock
OpenAI frontier models and Codex are now available on Amazon Bedrock, letting AWS customers use them inside existing cloud environments. OpenAI and AWS said the integration is designed to reduce procurement, security review, and deployment friction.
OpenAI frontier models and Codex are now available through Amazon Bedrock, bringing OpenAI capabilities into existing AWS environments. The companies said the move is meant to help customers evaluate and deploy AI systems faster while using the security and governance controls they already have in place.
Amazon Bedrock is AWS’s platform for building generative AI applications and agents at production scale. According to AWS, it gives teams access to AWS-native security, governance, and compliance controls, which can matter for organizations that need to keep AI workloads inside established cloud and policy boundaries.
⚡ New to this?
This news means companies that already run on AWS can now use some OpenAI models and developer tools without moving outside their current cloud setup. AWS, or Amazon Web Services, is a major cloud platform used for hosting apps, data, and infrastructure.
Amazon Bedrock is AWS’s service for building AI apps and agents. An AI agent is software that can carry out tasks with some independence, such as following multi-step workflows.
For non-experts, the big point is that OpenAI is trying to fit its tools into the controls businesses already use, which can make adoption easier for IT, security, and engineering teams.
🦞 OpenClaw angle
If you build self-hosted automations, treat this as a sign that cloud AI vendors are pushing harder on enterprise controls, not just model quality. Review which parts of your workflows depend on vendor-managed identity, governance, and orchestration, then decide whether those belong in your own stack or in a managed service like Bedrock.
If you run AI agents on AWS, map your current approval, logging, and compliance steps against what Bedrock already provides before adding extra tooling. For code assistants like Codex, test whether the service fits your existing dev environment and policy rules before wiring it into production pipelines.
If you are designing a hybrid setup, keep sensitive workflows and security automation separate from general productivity tasks so you can swap providers later without rebuilding everything.
The new availability includes OpenAI models on Bedrock and Bedrock Managed Agents powered by OpenAI. AWS said this combines OpenAI frontier models and agentic capabilities with AWS infrastructure, making it easier for organizations to build and deploy AI agents in the cloud.
OpenAI said the Bedrock path uses the OpenAI harness, which helps agents execute tasks faster, improve reasoning, and stay on task during long-running workflows. AWS said its infrastructure handles memory, identity, orchestration, security, and governance inside the customer’s environment, which can reduce operational overhead for teams moving from development to production.
Codex is also available on Amazon Bedrock. OpenAI said Codex helps developers write, review, debug, and modernize code in the tools and environments they already use to build and ship software.
The companies said customers can deploy OpenAI models through Amazon Bedrock within AWS environments while keeping existing security, governance, and compliance controls. OpenAI said this should reduce friction around procurement, security review, and production readiness.
“As customers begin using these capabilities, the AWS path helps reduce friction around procurement, security review, and production readiness,” OpenAI said. “By making OpenAI capabilities available within familiar AWS environments, organizations can spend less time navigating operational barriers and more time building.”
OpenAI and AWS said they plan to keep adding new capabilities to Amazon Bedrock. They also said they intend to bring specialized offerings such as Daybreak, OpenAI’s cybersecurity initiative, to AWS, so security teams can adopt AI-assisted software defense through their existing governance, procurement, security, and operational frameworks.
