Amazon Web Services says Canadian organizations can access OpenAI GPT-5.6, GPT-6, and GPT-6.1 models on Amazon Bedrock through cross-Region inference. The models are available from the Canada (Central) and Canada West (Calgary) AWS Regions, using inference profiles that route requests from a source Region to supported destination Regions.

The available models include GPT-5.6 Sol, Terra, and Luna; GPT-6 Astra, Sol, and Luna; and GPT-6.1 Sol. AWS describes the models as supporting text and image inputs, text outputs, a 1-million-token context window, reasoning mode, and tool calling. The post also says they support prompt caching on Amazon Bedrock.

Organizations can select a US geographic profile, which AWS says keeps processing within North America by using Canada and US Regions, or a global profile that routes requests to supported commercial AWS Regions. AWS notes that profile membership and model availability can change, and that the GPT-5.6 global profile is available from Canada (Central), but not Canada West.

The models can be accessed through the OpenAI Responses API, OpenAI Chat Completions API, and Amazon Bedrock Converse API on the Bedrock Runtime endpoint. AWS’s setup guidance covers IAM permissions, service control policies, profile verification, quota requests, and observability. The post also provides examples using Python, the OpenAI SDK, Boto3, and temporary Amazon Bedrock credentials.