← Back
AWS ML Blog

Accessing OpenAI models on Amazon Bedrock from Australia with global cross-Region inference

•11 min read•
#bedrock#llm#amazon
Level:Intermediate
For:AI Engineers
✦TL;DR

Australian teams can now invoke OpenAI’s GPT‑5.6 family (Sol, Terra, Luna) via Amazon Bedrock with global cross‑Region inference from the Asia Pacific (Sydney) ap‑southeast‑2 and Asia Pacific (Melbourne) ap‑southeast‑4 regions. All three models accept text and image inputs, support up to a 1 million‑token context window, and are reachable through the Bedrock Runtime’s Responses, Chat Completions, and Converse APIs. Prompt‑caching, OIDC authentication, and CloudWatch‑based monitoring are provided to reduce cost and observe usage. Engineers can call the models directly from their source region without handling destination routing, simplifying production deployments.

⚡ Key Takeaways

  • GPT‑5.6 Sol, Terra, and Luna are available with global inference profiles global.openai.gpt-5.6-sol, global.openai.gpt-5.6-terra, global.openai.gpt-5.6-luna.
  • Each model supports text + image inputs, generates text, and offers a 1 million‑token context window.
  • Cross‑Region routing lets a Sydney or Melbourne request be processed in any supported commercial AWS Region, expanding capacity without manual routing.
  • Invoke the models via the Bedrock Runtime using the Responses API, Chat Completions API, or Converse API.
  • Prerequisites: AWS account with Sydney or Melbourne enabled, IAM role/user with invoke permissions, SCP that allows the global profiles, Python 3.9+, and the openai, boto3, aws‑bedrock‑token‑generator packages.

🔧 Tools & Libraries

openaiboto3aws-bedrock-token-generatorAWS CLIAmazon Bedrock consoleAmazon CloudWatchCoding Agent Insights
💡 Why It Matters

This eliminates the need for engineers to manage multi‑region endpoint logic, giving instant access to the latest OpenAI models with built‑in cost‑optimisation (prompt caching) and observability, accelerating AI‑powered product rollouts in Australia.

✅ Practical Steps

  1. Install the required SDKs (`pip install openai boto3 aws-bedrock-token-generator`) and configure your IAM credentials.
  2. Verify active inference profiles with the AWS CLI, e.g.:
  3. `aws bedrock list-inference-profiles --region ap-southeast-2 --type-equals SYSTEM_DEFINED --query "inferenceProfileSummaries[?contains(inferenceProfileId, 'openai.gpt-5.6')].[inferenceProfileId,status]" --output table`
  4. Call a model from your Python code using the Bedrock Runtime endpoint for your source region (Sydney or Melbourne) and the desired API (Responses, Chat Completions, or Converse).

Want the full story? Read the original article.

Read on AWS ML Blog ↗

More like this

Censored vs. Truncated Regression in Python, Explained

Towards Data Science•#llm

How Google Duplex evolved to also call businesses for you in Gemini

Google Research•#agents

Q&A: Enabling more human-like interactions and understanding in multimodal foundation models

MIT News AI•#llm

Welcome Inkling 2 Mini, Cutting-edge Performance for your Local Setup

Hugging Face Blog•#llm

EXPLORE AI NEWS

Daily hand-picked stories on LLMs, RAG, agents and production AI — curated for engineers who ship.

BROWSE NEWS

GET THE WEEKLY DIGEST

Join engineers getting the Monday signal-over-noise AI breakdown. No spam, unsubscribe anytime.

LEARN AI ENGINEERING

Curated courses, research papers, repos and tutorials built for engineers leveling up in AI.

START LEARNING