India to develop own generative AI model: Ashwini Vaishnaw

Amid a tug of war on artificial intelligence (AI) supremacy between the US and China, union minister Ashwini Vaishnaw has announced that India will develop its own generative AI model.

The initiative was announced by the minister at the Utkarsh Odisha Conclave. He said that the model will be powered by the India AI Compute Facility, which has secured 19,000 GPUs to develop a large language model (LLM) designed specifically for the country. Out of these, 15,000 are high-end. With this, the government has overreached its target of 10,000 GPUs.

At least six developers have been identified who can roll this out in 8-10 months.

Foundation models are AI models trained on vast repositories of data, which can be used as the base for other applications.

“Cost of compute — GPU access per hour — after the 40% govt incentive will be less than a dollar in India, the lowest in the world,” the minister said in a press conference regarding the India AI Mission.

ET was the first to report on January 23 that India wants to build its own foundational model.

The Centre’s move follows the US government restricting export of AI compute infrastructure and limiting use of language models that are currently dominated by American technology firms. After the model is launched, this will place India in a select group of nations including the US and China where local companies are building prominent models.