OpenAI has announced local processing for its API in India, allowing eligible businesses and developers to run supported GPT-6 Luna requests within the country. The launch extends its existing data residency options beyond storage to include the processing that generates an AI response.
For Indian organisations evaluating AI tools, the distinction can be significant: storing information locally does not, by itself, mean the model processes that information in the same country. OpenAI says its new offering brings both processing and storage of supported customer content to India, with an initial focus on direct API usage.
What OpenAI's India Inference Residency Changes
Inference is the computation an AI model performs when it receives a request and generates a response. Under the new India offering, both the processing of a supported API request and the model computation behind its response take place in the country, according to OpenAI.
The company is targeting businesses whose deployment decisions depend on where information is processed. Financial services is one example highlighted in the announcement, although organisations will still need to assess whether the service meets their specific localisation and operational requirements.
Supported uses include text-based tasks such as summarising documents, drafting responses and helping employees analyse information. For a business building an internal document assistant, for example, the relevant change is the option to keep supported processing in India when using the eligible model and API configuration.
Supported Model and Availability
Local processing initially supports GPT-6 Luna input and output through the Responses and Chat Completions APIs. OpenAI says eligible customers can get started by creating a new API project, selecting India for both storage and processing, and using the corresponding India regional endpoint.
| Detail | Announced Support |
|---|---|
| Initial model | GPT-6 Luna |
| APIs | Responses and Chat Completions |
| Processing location | India, for supported requests |
| Customer content storage | India, for supported content |
| Setup | New API project with India selected for storage and processing |
The announcement covers direct API usage. It does not announce a change to where ordinary ChatGPT conversations are processed, or local processing for every OpenAI model. Businesses should therefore check the supported workloads and eligibility rather than assume the option applies across all their OpenAI usage.
Pricing, detailed eligibility criteria and a broader model rollout schedule were not specified in the media release.
Part of OpenAI's Wider India Expansion
OpenAI linked the launch to its OpenAI for India initiative, which spans infrastructure, enterprise adoption, skills and local partnerships. The company said that initiative included a partnership with Tata Group to develop domestic AI data centre capacity.
India is also part of a wider expansion of API inference residency that includes Australia and Japan. For Indian developers and enterprise customers, the immediate development is a local processing option for supported GPT-6 Luna workloads, giving organisations another factor to assess when deciding which AI applications they can deploy.
