OpenAI has launched OpenAI India data residency for its API, so supported requests are processed and computed in India and customer content is stored locally. The launch, announced on 12 November 2026, starts with GPT-6 Luna, which OpenAI describes as its most efficient model for focused, high-volume tasks.
Customers set it up by creating a new API project with India selected for storage and processing, then calling the endpoint in.api.openai.com. It works with the Responses and Chat Completions APIs and accepts text and image inputs with text output.
How does OpenAI’s India inference residency work?
Residency covers customer content, not system data, which may still be processed or stored outside India. Typical uses named for GPT-6 Luna include document summarisation, drafting responses and analysing information.
Access is not self-serve. Customers go through OpenAI’s sales team and a Modified Retention amendment, and they must have Zero Data Retention active. Modified Abuse Monitoring alone is not enough. Only Standard pay-as-you-go processing is supported, hosted tools are unavailable for in-region processing, and image inputs need separate approval for enhanced retention controls.
Data residency endpoints carry a 10% price uplift for eligible models released on or after 5 March 2026.
What it means for Indian businesses
Companies that want to keep customer content inside India can now run supported GPT-6 Luna workloads locally. The conditions above, including Zero Data Retention and the sales-led setup, mean it is aimed at enterprise users rather than casual developers.
In August 2026, Amazon Web Services added an India geographic inference profile on Amazon Bedrock for GPT-5.6 Terra and Luna. The new launch gives direct OpenAI customers a comparable option. The announcement is also linked to OpenAI’s wider India push, including its strategic partnership with Tata Group.
Industry Reaction and Expert Commentary
Nitin Bawankule, Head of Enterprise, India, at OpenAI, said: “Local inference gives them more control over where supported workloads run.” No independent industry reaction was found at the time of writing.
What Happens Next?
The source does not give a timeline for adding more models or features to India residency. Interested customers need to contact OpenAI’s sales team to be onboarded.
Frequently Asked Questions
What did OpenAI launch in India?
An inference residency for its API, so supported requests are processed in India and customer content is stored locally, starting with GPT-6 Luna.
Does it cost extra?
Yes. A 10% uplift applies to data residency endpoints for eligible models released on or after 5 March 2026.
Is all data kept in India?
No. It covers customer content, not system data, which may still be processed or stored outside India.
Source: Elets CIO, 12 November 2026.
Leave a comment