OpenAI, the American tech behemoth, stated that India is among the most developed AI marketplaces in the world, particularly in the areas of data analysis, coding, and higher-level reasoning.

In a statement about reasoning tokens that ChatGPT Plus users employ, OpenAI stated that India is among the top five countries in the world for per capita use of cognitive capacity.

With a fourfold increase in Codex users in just two weeks following the release of the company’s Codex app in February 2026, the nation has become one of the fastest-growing AI builder ecosystems.

According to the firm, India’s use of artificial intelligence (AI) is still heavily concentrated in a few cities, and its sophisticated use is even more unevenly distributed, indicating a significant opportunity to extend its use beyond early adopters in the top few cities.

“The key issue now is how rapidly the advantages of AI may spread from early adopters and major cities to the larger public. The young, quick-adopting Indian population will have a significant impact on the measures needed to close this gap, including increasing access, developing skills, and facilitating more meaningful use nationwide,” according to Oliver Jay, Managing Director-International, OpenAI.

As the top 10 cities only make up around 10% of the population but have around half of the AI users, advanced capability is heavily concentrated.

In India, Delhi-NCR has the greatest ChatGPT penetration rate. According to the study, India has a high level of AI competitiveness, but its greatest strengths are concentrated in a few cities, such as Bengaluru, Hyderabad, Delhi, and Chennai.

The Capability Gap results show that some of the most significant use cases for AI are already occurring in healthcare and education, among other fields, demonstrating how it is being used in real-world settings.

The firm discovered significant intra-regional disparities in sophisticated applications, with data analysis utilization being 30 times higher in top cities than in laggard cities.

The usage of coding was discovered to be four times greater, while the use of Codex developer was nine times higher.