What Exactly Has Changed?
OpenAI has enabled in-country processing for certain API requests, meaning that for supported services, the entire computational process—from receiving a prompt to generating a response—happens on servers located within India. This is an expansion of
its previous data residency offering, which only guaranteed that customer data was stored 'at rest' in the country. Now, both the processing (known as inference) and the storage of supported customer content will take place locally. This development is primarily aimed at businesses, especially those in regulated industries like finance, that have been hesitant to adopt AI due to strict data localisation mandates.
Which API Requests Are Supported?
At launch, the local processing capability is not universal across all of OpenAI's offerings. It specifically supports the GPT-6 Luna model. Developers can access this local processing through two primary endpoints: the Responses API and the Chat Completions API. These APIs are typically used for text-based tasks like document summarisation, drafting emails, and analysing information. To use this feature, developers must create a new API project and explicitly select India as the region for both processing and storage, then use the designated India regional endpoint. It's important to note that access may have eligibility requirements; some reports indicate the offering is for organisations with active Zero Data Retention (ZDR) agreements.
The Data Sovereignty Solution
For many Indian companies, the inability to process data within the country's borders has been a major roadblock to AI adoption. Sectors like banking, financial services, and insurance operate under regulations, such as those from the Reserve Bank of India, that often mandate local data handling for sensitive information. By bringing the entire inference process onshore, OpenAI directly addresses these data sovereignty concerns. Nitin Bawankule, OpenAI's Head of Enterprise for India, noted that this gives businesses more control and removes a 'practical barrier to adoption'. This move could unlock AI deployment for critical workflows that were previously considered too risky from a compliance perspective.
Lower Latency, Better Performance
Beyond compliance, local processing has a significant practical benefit: speed. When an API request is processed in a data centre thousands of miles away, the delay, or latency, can be noticeable. For applications that require real-time interaction, like chatbots or instant data analysis, every millisecond counts. By processing requests within India, the physical distance the data travels is dramatically reduced. This results in faster response times, creating a smoother and more responsive user experience. One early report cited a company that reduced latency from 1.1 seconds to under 300 milliseconds after switching to a local API. This performance boost makes a new class of real-time AI applications more viable for Indian developers and startups.
Part of a Broader India Strategy
This API launch is a key part of OpenAI's larger 'OpenAI for India' initiative, signaling a deep commitment to the market. This broader strategy includes developing local AI infrastructure, promoting enterprise adoption, and fostering skills development. Notably, OpenAI has been collaborating with Tata Group to build out domestic AI data centre capacity, which provides the physical foundation for this new local processing capability. By establishing a physical and compliant infrastructure footprint, OpenAI is positioning itself not just as a global service provider but as an integrated partner in India's rapidly growing digital ecosystem.
















