DeepSeek May Boost Race for India AI Model

Chinese firm's low-cost model raises hopes of democratizing AI globally

- Leslie D'Monte & Shouvik Das

DeepSeek AI, a Chinese AI startup, has rattled US tech stocks and global equity markets, even wiping out $593 billion from the value of AI posterboy Nvidia Corp. on Tuesday.

DeepSeek has released an open-source AI model that costs a fraction of OpenAI's and other large language models that are used to train AI, suggesting that AI models in countries like India do not need exorbitant high-end graphics processing units (GPUs) or gigantic data centres to be effective. Mint explains:

How did DeepSeek slash AI costs?

The DeepSeek-V3 model, launched on 10 January, used Nvidia's lower-capability H800 chips for training, spending less than $6 million, its researchers said in a paper. In contrast, OpenAI is said to have spent about $7 billion last year to train and operate ChatGPT, according to The Information, with an additional $3 billion for training the models with new data.