
Amazon Web Services (AWS) has made OpenAI’s GPT-5.6 Terra and GPT-5.6 Luna models available in India through Amazon Bedrock, giving Indian customers access to the models with India-based inference capabilities. The availability is aimed at organisations that need advanced AI capabilities while also meeting requirements around data residency and enterprise controls.
Through Amazon Bedrock, customers in India can use India Geo cross-Region inference, which routes requests only across AWS Regions within the India geography. The supported regions include Asia Pacific (Mumbai) and Asia Pacific (Hyderabad). This allows inference to be processed within India while providing the ability to scale workloads across AWS infrastructure.
The two models serve different workloads. GPT-5.6 Terra is positioned as a balanced model for applications that require a combination of intelligence, speed and cost efficiency, while GPT-5.6 Luna is designed for faster, high-volume and cost-sensitive workloads. AWS lists use cases such as content processing, classification, customer-service automation and other production workloads for the models.
The India availability also includes AWS controls for enterprise deployments. Customers can use features such as model invocation logging, Amazon CloudWatch monitoring and cost reporting through AWS Cost Explorer and the AWS Cost and Usage Report. The models support the Responses, Chat Completions and Converse APIs through Amazon Bedrock.
The move follows the broader availability of OpenAI’s GPT-5.6 family on Amazon Bedrock. AWS initially made GPT-5.6 Sol, Terra and Luna generally available on Bedrock in July 2026. Sol is positioned for more demanding reasoning workloads, while Terra and Luna target balanced and high-volume applications respectively.
Pricing for the GPT-5.6 models has also changed. OpenAI announced on July 30 that GPT-5.6 Luna would become 80% cheaper and GPT-5.6 Terra 20% cheaper, with the reductions reflected in Amazon Bedrock pricing as well.
For Indian enterprises, the combination of OpenAI models, AWS infrastructure and India-based inference provides an option for deploying AI workloads while addressing data-residency requirements. The availability expands the range of advanced AI models that organisations in India can access through Amazon Bedrock and its existing enterprise security, monitoring and governance controls.




