models
Amazon Bedrock Supports India Cross-Region Inference for GPT-5.6 Models
Amazon Bedrock now enables India-based cross-Region inference for OpenAI GPT-5.6 models, Terra and Luna, ensuring data remains within India while scaling AI workloads.
AS1 NewsSource: aws.amazon.com
Amazon Bedrock has introduced support for the OpenAI GPT-5.6 models, Terra and Luna, with geographic cross-Region inference capabilities within India. This feature allows users with local data processing needs, such as in financial services, healthcare, and public sectors, to utilize these models at scale while ensuring inference requests and data stay within Indian borders.
The cross-Region inference mechanism automatically routes requests between the Mumbai (ap-south-1) and Hyderabad (ap-south-2) AWS Regions, providing a capacity pool that helps maintain throughput and performance during traffic peaks. Users can initiate requests from either Region, with Amazon Bedrock managing routing based on capacity, thus supporting scalability and compliance with data residency requirements.
This functionality is implemented through inference profiles, which specify the model and routing regions. Two profiles are available for GPT-5.6 models: in.openai.gpt-5.6-terra and in.openai.gpt-5.6-luna. Billing and monitoring are centralized in the source Region, with logs recorded via Amazon CloudWatch and AWS CloudTrail.
The models support a 1-million-token context window, accept both text and image inputs, and produce text outputs, enabling processing of lengthy documents, large codebases, and mixed workloads. Developers can access GPT-5.6 models through the Amazon Bedrock console's text playground, the Responses API, Chat Completions API, or the Converse API, with options for multi-turn conversations, streaming responses, and prompt caching.
For organizations, appropriate IAM permissions are required to invoke models via the India geographic inference profiles, with policies scoped to the specific profile and model resources. The setup supports compliance with local data regulations, with data encrypted in transit and a zero data retention policy, except for flagged content used for abuse detection.
Overall, this update enhances AI deployment options within India, supporting scalable, compliant, and high-performance AI applications across various sectors.
Enables scalable, compliant deployment of GPT-5.6 models within India, supporting enterprise and government AI workloads.