Amid DeepSeek buzz, Ashwini Vaishnaw provides update on India’s own generative AI model

India is set to develop its own generative artificial intelligence (AI) model, joining the global AI race alongside major players like OpenAI’s ChatGPT and China’s DeepSeek, Union IT Minister Ashwini Vaishnaw said on Tuesday. The initiative will be powered by the India AI Compute Facility, which has secured 18,000 GPUs to support the development of a Large Language Model (LLM).

“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,” Vaishnaw said at the Utkarsh Odisha Conclave.

He also announced plans to establish AI data centres in Odisha.

In March 2024, the Union Cabinet approved IndiaAI mission with an outlay of ₹10,371.92 crore to establish a robust AI ecosystem by fostering public-private partnerships and providing critical infrastructure for AI innovation.

A key component of the mission is the IndiaAI Compute Facility, designed to build a scalable AI computing ecosystem for the country’s rapidly growing AI startups and research community. Under the mission, an AI marketplace will be established to offer AI as a service, along with pre-trained models for Indian innovators.

The IndiaAI Innovation Centre will spearhead the development of indigenous Large Multimodal Models (LMMs) and domain-specific foundational models, catering to critical sectors such as healthcare, agriculture, and governance. A dedicated IndiaAI Datasets Platform will provide seamless access to high-quality, non-personal datasets, enabling startups and researchers to drive AI-driven solutions.

The thirteen shortlisted companies were reportedly selected from a total of 19 bidders. These include Jio Platforms, Tata Communications, CtrlS Datacenters Limited, E2E Networks, Yotta Data Services, CMS Computers, Hostin Services, I2k2 Networks, Orient Technologies, Vensysco Technologies, CloudThat Technologies, Locuz Enterprise Solutions, and NxtGen Datacenter and Cloud Technologies.

China's rapid progress on AI front

The announcement regarding India’s generative AI model comes amid the emergence of China’s DeepSeek, which has sent ripples through the global AI industry.

The Chinese firm claims its DeepSeek-R1 model operates at 95% lower costs than comparable OpenAI models while achieving a 97% success rate in certain tasks. The development has raised questions about the long-term dominance of US AI giants, including OpenAI and Google.

Following DeepSeek’s rise, major AI-linked stocks such as Microsoft, Nvidia Corp., Oracle Corp., and Google parent Alphabet Inc. saw a sharp sell-off on Monday, wiping out nearly $1 trillion in combined market value.

Meanwhile, Chinese tech giant Alibaba on Wednesday released Qwen 2.5, an upgraded AI model that it claims outperforms DeepSeek-V3. Alibaba’s cloud division stated that Qwen 2.5-Max surpasses OpenAI’s GPT-4o, DeepSeek-V3, and Meta’s Llama-3.1-405B across multiple AI performance benchmarks.

DeepSeek’s rapid ascent has not only disrupted global AI markets but also prompted domestic competitors, including ByteDance, to accelerate their own AI advancements.