Access OpenAI GPT models on Amazon Bedrock from Canada with cross-Region inference
Amazon Web Services Roopesh Ramklass
OpenAI GPT-5.6, GPT-6, and GPT-6.1 are now reachable on Amazon Bedrock from Canada. AWS routes the request for you, so Canadian apps get more capacity without rewriting routing.
Based on reporting by Amazon Web Services, Roopesh Ramklass — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Canadian teams building on Amazon Bedrock can now reach OpenAI’s GPT-5.6, GPT-6, and GPT-6.1 from Canada’s AWS Regions through cross-Region inference. That means apps in Canada (Central) and Canada West (Calgary) can call the models locally, while AWS handles where the request is actually processed.
The lineup is broader than a single model drop. GPT-6.1 Sol is aimed at strong reasoning and coding where performance and cost both matter. GPT-6 comes in Astra, Sol, and Luna variants, with Astra positioned for complex end-to-end reasoning and agentic work, Sol for recurring reasoning and coding tasks, and Luna for fast, low-cost inference on repeated workloads. GPT-5.6 has Sol, Terra, and Luna versions for high-reasoning work, balanced everyday production, and high-volume, latency-sensitive use.
These models accept text and image inputs, return text, and support a 1-million-token context window, reasoning mode, and tool calling. GPT-5.6, GPT-6, and GPT-6.1 also support prompt caching on Amazon Bedrock. The models are available through the OpenAI Responses API, the OpenAI Chat Completions API, and the Amazon Bedrock Converse API on the bedrock-runtime endpoint.
AWS is using inference profiles as the control plane here. A profile is a logical ID that tells Bedrock which model to use and which Regions it can route to. From Canada, there are two kinds: US geographic profiles, which keep inference within North America, and global profiles, which can use supported commercial AWS Regions where the model is deployed. For GPT-5.6, the global profile is available only from ca-central-1; from ca-west-1, AWS says to use the US geographic profile.
The setup is straightforward if your IAM permissions and SCPs are already in order. AWS also shows how to verify active profiles with the CLI or console, invoke the models through the OpenAI SDK or boto3, stream Responses API output, use tool calls, and track quota use in tokens per minute. The practical message is simple: Canadian customers get the new OpenAI models without giving up Bedrock’s routing, caching, or access controls.
My take — AI-written commentary, not fact-checked reporting
This is the kind of cloud feature that matters more than the headline model names. Canadian data-boundary worries are real, and AWS is clearly trying to make “we can use the new thing” compatible with “we can still sleep at night.” The odd part is how often AI infrastructure still feels like paperwork with better branding.
Read more about this at: Amazon Web Services