OpenAI GPT-5.6, GPT-6, GPT-6.1 on Bedrock from Canada (Central)

The availability builds on Amazon Bedrock's existing cross-Region inference mechanism, which lets applications call a local endpoint while Bedrock routes requests to a broader capacity pool. From Canada, two profile types are offered: a US geographic profile (prefixed us.) that keeps processing within North America (Canada and the US), and a global profile (prefixed global.) that draws on a wider set of commercial AWS Regions. The choice matters for workloads with data-boundary requirements, and the body notes that the global profile for GPT-5.6 models is available only from ca-central-1, while ca-west-1 users must use the US geographic profile for GPT-5.6.

The models accept text and image inputs, return text, and support a 1-million-token context window, reasoning mode, and tool calling. They are accessible through the OpenAI Responses API, the OpenAI Chat Completions API, and the Amazon Bedrock Converse API on the bedrock-runtime endpoint. For cost control, prompt caching automatically caches eligible prefixes of 1,024 tokens or more, and cached tokens are billed at a discounted rate. Quotas are measured in tokens per minute per model, with separate allocations for US geographic and global profiles, so switching profiles changes the capacity pool.

Security and compliance details are also provided: requests travel over the AWS private backbone encrypted with TLS 1.2+, data at rest stays in the source Region by default, and model providers do not have access to prompts or completions by default. One OpenAI-specific consideration is that content flagged by automated abuse classifiers might be retained for up to 30 days in the destination Region for review, which the body recommends reviewing with a compliance team.