OpenAI models and Codex are now generally available on Amazon Bedrock
OpenAI frontier models and Codex are now generally available through Amazon Bedrock, giving AWS teams a production path that uses existing IAM, billing, governance, and deployment controls.
OpenAI frontier models and Codex are now generally available on Amazon Bedrock as of June 1, 2026. AWS says teams can use GPT-5.5, GPT-5.4, and Codex through Bedrock, with OpenAI-compatible API paths and AWS-native controls. The practical change is procurement and deployment: teams already standardized on AWS can evaluate OpenAI models and coding-agent workflows without moving the production control plane outside their AWS environment.
Key takeaways
- AWS says GPT-5.5, GPT-5.4, and Codex are now generally available on Amazon Bedrock.
- The AWS post says pricing matches OpenAI first-party per-token rates and has no additional Bedrock markup for these models.
- OpenAI says the AWS path is meant for existing security, compliance, procurement, billing, and governance workflows.
- Codex on Bedrock targets software engineering work such as writing, reviewing, debugging, and modernization inside AWS-managed environments.
- Daybreak and Codex Security are described as future AWS availability, not something to treat as live in this rollout.
Practical LinkLoot angle
This is a routing and governance decision more than a model announcement. If your team already runs workloads in AWS, Bedrock can become the controlled entry point for OpenAI inference, model evaluation, and Codex adoption. That matters for teams that need IAM, CloudTrail, procurement approval, committed-spend accounting, or region-specific deployment before they can move an AI workflow from a prototype into production.
| Path | Best use | Limitation to check | Source |
|---|---|---|---|
| OpenAI models on Bedrock | Production apps that need OpenAI models inside AWS controls | Confirm model and region availability before committing architecture | AWS / OpenAI |
| Codex on Bedrock | Coding-agent workflows tied to AWS governance and billing | Verify supported Codex clients and your team's permission model | AWS / OpenAI |
| Direct OpenAI API | Fastest path when AWS governance is not a blocker | Separate procurement, monitoring, and compliance controls may still be needed | OpenAI |
| Future Daybreak availability | Security review, threat modeling, and remediation workflows later | OpenAI describes this as future availability, not current GA | OpenAI |
For builders comparing agent stacks, pair this with LinkLoot's AI agent tools guide and evaluate where the agent runs, what account owns the logs, and who can approve tool access.
What to verify before you act
Check the exact Bedrock region list and model IDs your account can access. AWS availability can differ by region, account status, quota, and service rollout phase, so a blog announcement is not a substitute for a console or API check in your own environment.
Confirm the cost model before routing agentic workflows through Bedrock. AWS says pricing matches OpenAI first-party rates for these models, but your internal accounting may still treat Bedrock usage differently because it can count toward AWS commitments and existing cloud budgets.
Review Codex governance separately from normal model inference. A coding agent can read code, propose changes, run tools, and touch development workflows, so permissions, repository boundaries, audit logs, and approval steps matter more than they do for a simple chat completion endpoint.
Source check
AWS confirms general availability for GPT-5.5, GPT-5.4, and Codex on Amazon Bedrock, plus the stated pricing position and production deployment framing. OpenAI confirms the AWS availability and positions it around existing enterprise governance, procurement, security, and deployment workflows. Help Net Security independently reports the same availability and highlights the enterprise security and governance angle.
Yes. AWS and OpenAI say OpenAI frontier models and Codex are generally available on Amazon Bedrock as of June 1, 2026.