In a significant endorsement, IT Minister Ashwini Vaishnaw praised Chinese startup DeepSeek for its innovative low-cost AI assistant, drawing parallels between the company’s frugal approach and the Indian government’s initiatives to develop a localized AI model.
This comes at a time when India has pledged to invest $1.25 billion in artificial intelligence, announced in March, under the IndiaAI mission, which aims to fund AI startups and establish a robust AI infrastructure.
Commenting on an event in Odisha, Vaishnaw said, “Some people question how much investments the government has made to the IndiaAI mission. Look at what DeepSeek has achieved: developing a powerful model for just $5.5 million. It’s all about effective utilization of intelligence.”
The rapid rise of the firm has led to a global reassessment of AI budgeting. DeepSeek said it only spent two months and less than $6 million of development money creating its AI model, using Nvidia’s H800 chips-which many consider last-generation models. During the most recent period, the app from the startup has surpassed OpenAI’s ChatGPT in downloads from Apple’s App Store-which was the last bastion of insight, as many questions how China is lagging behind the U.S. in AI-related developments.
Vaishnaw’s comments would appear to respond to the remarks made by OpenAI’s CEO Sam Altman when he visited India last year. Altman had reportedly doubted the ability of Indian teams to build large-scale models within the $10 million budget. He termed the idea of competing in training foundational models “totally hopeless.
Altman’s past quotes have also emerged on X in light of the success DeepSeek has garnered recently. He will be heading back to India on February 5, coinciding with the on-going legal cases his company has against the country’s digital news and book publishing houses on matters of copyright infringement.
