Amazon Web Services has announced the general availability of OpenAI's GPT-5.6 Terra and GPT-5.6 Luna models on Amazon Bedrock in India, enabling organisations to process model inference on AWS infrastructure within the country. According to reports from Amazon, this move is expected to benefit organisations across financial services, healthcare, public services and start-ups, allowing them to deploy OpenAI models locally while maintaining existing AWS security, governance and compliance controls. The launch enables banks, hospitals, and government offices to tap into advanced AI while keeping data secure and meeting local rules through the unified API integration. As reported by CNBC-TV18, OpenAI India's Head of Enterprise Sales Nitin Bawankule emphasized that "for many organisations in India, especially in regulated sectors, local inferencing is a key part of adopting AI with confidence."
The availability comes alongside recent pricing reductions that significantly lower deployment costs. As reported by AWS, GPT-5.6 Luna costs up to 80% less, while GPT-5.6 Terra costs up to 20% less following OpenAI's latest price changes. The reduced pricing is reflected on Amazon Bedrock, potentially lowering the cost of deploying AI applications at scale for Indian enterprises. The 80% cost reduction for Luna makes powerful AI a lot more accessible for businesses in India, significantly expanding the reach of advanced AI capabilities across various sectors. According to AWS, the rollout includes OpenAI's recently announced pricing reductions, making the models more affordable for Indian enterprises.
According to the release by Amazon, GPT-5.6 Terra is positioned as a balanced, general-purpose model for enterprise workloads, including software development, coding, data analysis and multi-step agent workflows. GPT-5.6 Luna is designed for high-volume and latency-sensitive applications such as summarisation, classification, customer support and voice assistants. The models can be used for agentic coding, data analysis, regulated-process automation, KYC verification, regulatory filings and document processing, with businesses also using Luna for high-volume customer-facing applications including AI-powered chatbots and voice assistants. Terra is built for everyday business tasks, while Luna is tuned for fast, high-volume jobs like chatbots or voice assistants. Potential enterprise use cases include code generation, review and testing; regulatory reporting and filings; Know-Your-Customer (KYC) verification; document processing and automation; AI-powered customer service systems; data analytics and workflow automation.
Through Amazon Bedrock's unified API, organisations can access OpenAI models alongside other foundation models available on AWS without having to substantially alter their existing infrastructure or development workflows. As reported by AWS, customers can continue using their existing security, governance and compliance mechanisms when deploying the OpenAI models through Bedrock. In-country inference could help organisations with data-residency and regulatory requirements where processing data within India is important, providing a combination of local inference, enterprise controls and reduced pricing for Indian businesses. Integration is easy: just one API gets you started without overhauling your current setup and allows teams to match each workload to the most cost-effective tier – routing everyday tasks to Terra and high-volume jobs to Luna. Every call to OpenAI models on Amazon Bedrock benefits from the enterprise-grade controls customers already use across AWS, enabling organisations to apply existing governance, security and operational practices as they bring advanced AI into production environments.
