New Delhi plans to launch an AI model by the end of 2025 to compete with DeepSeek and ChatGPT, Ashwini Vaishnaw has announced
India will develop its own large language model powered by artificial intelligence (AI) to compete with DeepSeek and ChatGPT, Minister of Electronics and IT Ashwini Vaishnaw told media on Thursday. It is expected to be “ready within the next 10 months,” he stated, adding that its foundational framework is already in place.
With the groundwork complete, the government is now focusing on creating a system tailored to India’s unique requirements, the minister noted.
As part of its artificial intelligence push, the government of Prime Minister Narendra Modi has approved 18 proposals aimed at accelerating AI solutions in key sectors such as agriculture and climate change. According to a Mint report, this support includes access to computing power, data, and funding.
Vaishnaw also revealed that six major developers are set to launch foundational AI models by the end of the year. To make related development more “affordable and accessible,” the minister said the government will subsidize 40% of computing costs, lowering the average price per AI computing unit from $1.29 per hour.
This initiative is a key component of the $1.2 billion IndiaAI mission, which seeks to develop both large and small language models. To strengthen domestic AI capabilities, New Delhi is working on building a computing infrastructure of over 18,000 graphics processing units (GPUs). Billionaire Mukesh Ambani’s Jio Platforms is among the companies racing to achieve this, utilizing advanced processors like Nvidia’s H100 chips.
Earlier this month, Bloomberg repotted that Ambani is planning to build what could become the world’s largest data center in Jamnagar, in the state of Gujarat. Ambani-led conglomerate Reliance Industries operates a massive refinery, considered world’s largest, in Jamnagar.
The data center is expected have a total capacity of three gigawatts, which would put India on the map in terms of advanced technological capabilities. Today’s largest operational data centers are mostly located in the US and are under one gigawatt.
Vaishnaw’s statements came shortly after the AI model DeepSeek gained global attention. It recently surpassed US-based OpenAI’s ChatGPT as the most popular AI assistant on Apple’s App Store. Analysts view the Chinese model’s breakthrough as evidence that AI innovation does not necessarily require massive capital investments, signaling a shift in how this kind of technological progress can be achieved globally.
