Set up OpenAI ChatGPT Codex with LiteLLM on Amazon ECS and Amazon Bedrock

AWS details how to run a customer-operated LiteLLM gateway on Amazon ECS with Fargate, link it to an OpenAI model on Bedrock, and route Codex requests with scoped identities, budgets, and telemetry.

By AIWI EditorialReviewed by Chayanika DivekarAI-assisted summary · human-reviewed1 min readPublished Sep 3, 2026

Single source1 source

Reality Check

Availability: Shipping now
Hype-o-meter: Mostly substance
Who it's for: Developers · Enterprises
Do this now: Review the described ECS/Fargate and Bedrock configuration to assess whether a LiteLLM gateway fits your Codex deployment.
Claims check: Input describes a concrete deployment pattern and comparison; no performance claims made.
Corroboration: Single source · 1 source

Key Takeaways

Deploy a LiteLLM gateway on Amazon ECS with AWS Fargate
Connect the gateway to an OpenAI model on Amazon Bedrock
Configure Codex to route through the gateway's Responses API
Compares direct IAM Identity Center access and a managed Portkey deployment

AWS describes how to set up OpenAI ChatGPT Codex with a LiteLLM gateway running on Amazon ECS using AWS Fargate. The gateway, operated by the customer, connects to an OpenAI model available through Amazon Bedrock. Codex is configured to route its requests through the gateway’s Responses API, with controls including scoped identities, budgets, rate limits, and telemetry. The post also compares alternative approaches: direct access via IAM Identity Center and a managed Portkey deployment.

Who should care

Developers and engineering teams working within AWS environments who want to route Codex traffic through a managed gateway with identity, budget, and rate-limit controls will find this setup relevant, along with those evaluating gateway options such as LiteLLM versus Portkey.

Was this helpful?