OpenAI announced on Thursday that it is now processing API requests locally in India, allowing eligible businesses and developers to run supported workloads inside the country rather than sending them to data centres abroad.
For the moment, local processing covers the GPT-6 Luna model's input and output through OpenAI's Responses and Chat Completions APIs. The company framed the launch as an expansion of data-residency options for API users, a category that has become a procurement requirement for regulated industries and for enterprises handling sensitive personal data.
Data residency — the assurance that data is stored and processed within a given jurisdiction — has become one of the sharpest points of friction between global AI providers and national regulators. India has pushed for local handling of data in several sectors, and cloud providers have responded with in-country regions and residency commitments. Announcing in-country inference, rather than only in-country storage, is the stronger claim: it means the model's computation itself stays inside the border.
OpenAI did not say how many customers are eligible, or whether the capability will extend to other models; the company said the launch currently supports direct API usage only. The announcement was reported by Indian outlets including Mint and the Press Trust of India.
The timing is notable. Enterprises weighing AI vendors increasingly ask where inference happens, not just where data sits at rest, and every provider that can answer "inside your borders" gains an advantage in markets where cross-border data transfer is politically sensitive.
Sources
- livemint.comOpenAI goes local in India, enables in-country GPT-6 Luna API inference (Mint)
- enterpriseai.economictimes.indiatimes.comOpenAI brings local AI processing to India, expands data residency options for API users (ET Enterprise AI)
- rediff.comOpenAI to process AI requests within India for businesses (Rediff / PTI)

