India plans own GenAI model as US-China rivalry deepens

Getting your Trinity Audio player ready...

India is preparing to create its own generative artificial intelligence (GenAI) model, Union IT Minister Ashwini Vaishnaw announced at the Utkarsh Odisha Conclave. This will position the South Asian nation alongside global AI leaders such as ChatGPT and DeepSeek.

The project will be supported by the India AI Compute Facility, which has obtained over 18,000 graphics processing unites (GPUs) to support the creation of a large language model (LLM) designed to meet the specific needs of the country.

LLMs serve as the core for GenAI. While global, general-purpose LLMs support a variety of use cases and have broad applicability, there are various reasons and challenges associated with developing models tailored to local languages.

“The entire world believes that India will be a major semiconductor destination…very soon we will have our own LLM,” Vaishnaw said. 

“With 18,000 GPUs in place, India is well on its way to building a homegrown AI model that will cater to the unique linguistic, economic, and social requirements of the nation,” he added.

The government has reportedly chosen 18 proposals to receive support in the form of computing infrastructure, data, and funding for the development of AI applications in areas like agriculture and climate change. According to Vaishnaw, India will cover 40% of the computing costs for these proposals.

“The foundational models made in India will be able to compete with the best of the best in the world,” Vaishnaw said.

“We believe that there are at least six major developers who can develop AI models in the six to eight months on the outer limit, and four to six months on a more optimistic estimate. A common compute facility is the most important component for creating a robust AI ecosystem,” Vaishnaw pointed out.

As a key element of the IndiaAI Mission, the government has focused on creating a centralized computing resource. Vaishnaw shared that the facility has surpassed initial goals, acquiring more than 18,000 GPUs, far exceeding the original target of 10,000. This reportedly includes 12,896 Nvidia (NASDAQ: NVDA) H100 GPUs and 1,480 Nvidia H200 GPUs, some of the most advanced AI processors on the market. Around 10,000 GPUs are already operational, with the remaining ones set to be rolled out over time.

Vaishnaw stated that AI compute services are being offered at an average discount of 42% compared to market prices, with high-precision compute units, essential for foundational models, available at a 47% discount.

India is also set to introduce its own foundational AI model, aimed at tackling the nation’s linguistic and cultural diversity while reducing biases in datasets. Vaishnaw revealed that calls for proposals to develop the AI model would begin with an anticipated development period of four to eight months.

Who will dominate the AI sector?

India’s announcement comes at a time when the United States and China are engaged in a fierce competition for dominance in AI. This move coincides with China’s DeepSeek AI startup making headlines with its breakthroughs, challenging U.S. industry giants like OpenAI.

DeepSeek, a Chinese AI startup barely a year old, has caused both excitement and concern in Silicon Valley by showcasing AI models that deliver performance on par with the world’s top chatbots, reportedly at a fraction of the development cost. On January 27, global tech stocks saw a decline as the buzz surrounding DeepSeek’s innovation grew, prompting investors to consider the potential impact on U.S. competitors and AI hardware providers like Nvidia Corp. U.S. chipmaker Nvidia, whose semiconductors are crucial to the AI industry, spearheaded a steep decline in tech stocks, shedding almost $600 billion in market value.

Nvidia’s stock suffered a 17% plunge, the largest single-day drop in market history. Other technology sector stocks, including those in semiconductor, power, and AI-related infrastructure sectors, suffered a collective loss of over $1 trillion in market value.

“The release of DeepSeek AI from a Chinese company should be a wake-up call for our industries that we need to be laser-focussed on competing to win because we have the greatest scientists in the world,” Donald Trump said, reacting to the bloodbath on Wall Street.

The statement comes just days after Trump, Oracle CEO Larry Ellison, SoftBank (NASDAQ: SFTBY) CEO Masayoshi Son, and OpenAI CEO Sam Altman announced “Stargate,” an AI data center venture. Stargate is a joint effort led by Oracle, SoftBank, and OpenAI, which aims to invest $100 billion to $500 billion in developing AI infrastructure in the United States.

On the other hand, billionaire Mukesh Ambani’s Reliance Group is buying Nvidia’s AI semiconductors to set up a data center in India. The data center is anticipated to have a total capacity of three gigawatts, the world’s largest data center by capacity.

DeepSeek on Indian servers

Vaishnaw announced that DeepSeek will soon be hosted on servers in India to tackle privacy concerns.

“It is an open-source model. Like LLama, which is open source, this too can be hosted on Indian servers. Data privacy issues regarding DeepSeek can be addressed by hosting open source models on Indian servers,” Vaishnaw explained. 

In 2024, India approved roughly $1.24 billion for the IndiaAI Mission, aimed at boosting the country’s AI ecosystem, innovation, and entrepreneurship.

“Some people question the amount of investment the government has committed to the (IndiaAI mission). You have seen what DeepSeek has done? $5.5 million and a very, very powerful model. Because of the use of the brain,” Vaishnaw said.

India is also planning to manufacture its own ‘Made in India’ chip by the end of 2025, as well as exploring the possibility of having its own GPUs in three to five years. While manufacturing its own chips is a significant move toward accelerating its semiconductor ambitions, India’s GPUs will help the country boost the development of AI infrastructure instead of solely relying on foreign suppliers.

In order for artificial intelligence (AI) to work right within the law and thrive in the face of growing challenges, it needs to integrate an enterprise blockchain system that ensures data input quality and ownership—allowing it to keep data safe while also guaranteeing the immutability of data. Check out CoinGeek’s coverage on this emerging tech to learn more why Enterprise blockchain will be the backbone of AI.

Watch: Blockchain, IPv6, AI & 5G will pave the way for the new Internet