OpenAI is making its latest AI models more accessible to Indian businesses through a new partnership with Amazon Web Services (AWS). The collaboration brings GPT-5.6 Terra and GPT-5.6 Luna to Amazon Bedrock in India, with a key advantage: AI processing can now take place within the country.
In practical terms, Indian businesses can access OpenAI’s advanced models through AWS while their requests are processed on AWS infrastructure located in India. This could be especially valuable for sectors such as banking, healthcare and government, where organisations handle sensitive information and face strict data-handling requirements.
The two models target different workloads. GPT-5.6 Terra is positioned as the more balanced model and is intended for common business requirements such as software development, information analysis and document-related tasks. GPT-5.6 Luna, on the other hand, is optimised for speed and cost efficiency, making it better suited to organisations that need to handle very large volumes of requests.
AWS says both models are now generally available on Amazon Bedrock in India. The company also notes that following OpenAI’s latest price reductions, GPT-5.6 Luna can cost up to 80 per cent less, while GPT-5.6 Terra can cost up to 20 per cent less.
Why bringing OpenAI models to India matters
The most significant aspect of the announcement is not simply that Indian businesses can access OpenAI’s models, but that the processing can happen within India.
Whenever someone sends a request to an AI model, computers running that model have to process the information. For organisations dealing with confidential or sensitive data, the location where that processing occurs can be extremely important. With in-country inferencing, AWS says requests made through OpenAI models on Amazon Bedrock are processed using AWS infrastructure in India.
This gives Indian businesses another way to deploy advanced AI while continuing to use the security, governance and operational controls they already have in place through AWS.
Amazon Bedrock also provides access to several AI models through a common API. In simpler terms, businesses do not necessarily need to completely redesign their AI applications whenever they want to experiment with or switch between models. They can select the model that best fits a particular workload.
For instance, a company could use Terra for routine business operations and choose Luna when it needs to handle substantially more requests while keeping costs down.
Potential applications span a wide range of business activities. Indian companies could use the models for software development and review, analysing large datasets, generating reports, automating document-intensive workflows and creating AI agents capable of completing multi-step tasks. Banks and healthcare organisations could apply them to processes such as KYC verification, document handling and regulatory workflows, while companies serving millions of customers could use Luna for chatbots and other high-volume applications.
For most consumers, the announcement is unlikely to immediately change how they interact with ChatGPT. The partnership is primarily focused on enterprise use. However, it could prove important for Indian companies looking to move beyond experimenting with AI and begin deploying it at scale.
