AWS Adds In-Country GPT-5.6 Inference for India on Amazon Bedrock

AWS has expanded Amazon Bedrock with India geographic cross-Region inference for OpenAI GPT-5.6 Terra and Luna. The important change is not simply another regional endpoint: the new in. inference profiles are designed so inference processing stays inside India, routing only between the Mumbai (ap-south-1) and Hyderabad (ap-south-2) Regions.

That makes the launch particularly relevant to teams with local processing requirements in regulated or data-sensitive sectors. AWS explicitly points to financial services, healthcare and the public sector as examples where keeping inference within the country can matter. The profiles are in.openai.gpt-5.6-terra and in.openai.gpt-5.6-luna.

Both models support text and image input, text output and a one-million-token context window. Developers can call them through Bedrock Runtime using Bedrock-native APIs as well as OpenAI-compatible Responses and Chat Completions interfaces. For teams already using OpenAI SDK patterns, this lowers the integration cost because the application can target Bedrock's OpenAI-compatible endpoint and use the India profile ID as the model identifier.

The operational boundary is also clear: local India profiles differ from global cross-Region inference. Global profiles can route to supported commercial AWS Regions worldwide for maximum capacity, while the India-prefixed profiles are the choice when processing must remain in-country. AWS records invocation metrics and audit information in the source Region, and CloudTrail exposes which India Region actually processed a request.

For enterprise architecture, the practical takeaway is that model selection is increasingly becoming a deployment-governance decision as much as a capability decision. Teams can now choose GPT-5.6 Terra or Luna on Bedrock while keeping inference routing inside India, then apply existing IAM, CloudWatch, CloudTrail and Bedrock controls around the workload.