Introducing India cross-Region inference for OpenAI GPT-5.6 models on Amazon Bedrock

This article announces India geographic cross-Region inference support for OpenAI GPT-5.6 models (Terra and Luna) on Amazon Bedrock, enabling compliant local data processing for financial services, healthcare, and public sector customers.

• GPT-5.6 models feature 1-million-token context window, accept text and image input, produce text output
• Cross-Region inference automatically routes requests between Mumbai (ap-south-1) and Hyderabad (ap-south-2) regions to improve throughput and maintain performance under load
• Data residency guaranteed: inference requests and data stay within India; encrypted in transit across AWS network
• Access via OpenAI Responses API, Chat Completions API, or Amazon Bedrock Converse API with inference profile IDs (in.openai.gpt-5.6-terra, in.openai.gpt-5.6-luna)
• Supports prompt caching with 90% discount on cached token reads, multi-turn conversations with server-side state, and streaming responses
• Billing and quota tracked in source region; CloudWatch metrics and CloudTrail logs recorded in source region only

India geographic inference profiles enable scalable deployment of advanced OpenAI models while maintaining strict data residency compliance for regulated industries.