Introducing OpenAI models on Amazon Bedrock for in-country inferencing in India
aws ·
Amazon Bedrock now supports OpenAI GPT-5.6 Terra and Luna in India through India geographic cross-Region inference, enabling workloads with local data-processing requirements to run at scale. Both models accept text and image inputs, generate text, and provide a 1-million-token context window for long documents, large codebases, and multimodal requests. Requests are routed between the Mumbai and Hyderabad AWS Regions using the profiles in.openai.gpt-5.6-terra and in.openai.gpt-5.6-luna, while billing, quotas, and monitoring records remain associated with the source Region. The India profiles keep inference within the country, with data encrypted in transit and Amazon Bedrock generally using zero data retention, although content flagged by abuse-detection systems may be retained for offline detection. Global cross-Region profiles are also available and can route requests to supported AWS Regions worldwide, but the article recommends India profiles for data-residency needs. Developers can use the Bedrock console, OpenAI Responses and Chat Completions APIs, or the Bedrock Converse API with AWS credentials or Bedrock API keys. The models also support configurable reasoning effort, server-side multi-turn state, streaming, and prompt caching, with cached reads billed at a 90 percent discount compared with uncached input tokens.
aws