What is DeepSeek, how did it shake up the US AI giants?
DeepSeek is a Chinese company of Artificial Intelligence (AI). It has developed a cheaper AI chip than American AI companies. In comparison to the way American companies developed AI models by investing huge amounts of money, DeepSeek has challenged them by developing a cheaper chip. DeepSeek is a private company in China, which was founded in July 2023 by Liang Wenfeng, a graduate of Zhejiang University, one of the largest universities in China. According to MIT Technology Review, Liang Wenfeng funded the DeepSeek startup through his hedge fund. Now Liang has assets worth about $8 billion.
DeepSeek’s stir started in America when it launched its brand new AI model. In America, this model was described as better and cheaper than the models of the AI giants there. It has cost so little that the American giants are not believing it. This is also the reason for their trouble. At present, American media outlets are talking the most about DeepSeek.
A paper from DeepSeek recently caught the attention of AI giants. In that paper, the company said that DeepSeek-v3 training requires less than $6 million of computing power and can be trained using Nvidia H800 chips.

Liang Wenfeng, Founder of DeepSeek
DeepSeek’s AI assistant app has overtaken its rival ChatGPT on Apple’s App Store in the US. This app is powered by DeepSeek-V3. It is now the top-rated free application on Apple’s App Store.
DeepSeek has raised doubts about the logic behind the decision of some US tech companies to spend billions of dollars in AI investments. Shares of several major AI players including US OpenAI and Nvidia have been directly affected.
The release of OpenAI’s ChatGPT in late 2022 created a stir among Chinese tech companies. Then China started research and other work on creating its own AI-powered chatbots. Chinese search engine giant Baidu released the first Chinese AI model equal to ChatGPT. But Baidu’s AI model was not good compared to ChatGPT. Baidu’s AI model was not strong compared to the AI capabilities of US companies. There was widespread disappointment in China. But now the quality and cost of DeepSeek’s models have turned this story upside down.
The Chinese startup said the two models, DeepSeek-V3 and DeepSeek-R1, which have been praised by Silicon Valley executives and US tech company engineers alike, are on par with the most advanced models from OpenAI and Meta. They are also cheaper to use. According to a post on DeepSeek’s official WeChat account, DeepSeek-R1, released last week, is 20 to 50 times cheaper to use than the OpenAI O1 model, depending on the task.
DeepSeek’s success is also being watched closely in China’s top political circles. According to China’s Xinhua news agency, on January 20, the day DeepSeek-R1 was released to the public, founder Liang attended a closed-door meeting with Chinese Prime Minister Li Keqiang, attended by other Chinese business and high-level officials.
Liang’s presence at this meeting is an indication that Beijing is also going to take some steps to overcome US export controls and achieve self-reliance in strategic industries like AI.