Today, according to The Information reported that two people familiar with the matter revealed that DeepSeek founder and CEO Liang Wenfeng may personally invest up to 20 billion yuan in the company’s first round of financing since its establishment, accounting for 40% of the total funds raised, becoming the largest investor in this round of financing .
DeepSeek hopes to raise 50 billion yuan in this round of financing, which will become the largest round of financing in the history of Chinese AI companies. According to a person familiar with the matter, China’s National Artificial Intelligence Industry Investment Fund (Big Fund) is expected to become DeepSeek’s second-largest investor in this round of financing.
At the same time, this round of financing also prompted DeepSeek to accelerate revenue generation and commercialization. Two people involved in the negotiations revealed that DeepSeek recently stated to some investors that it plans to speed up the pace of model release to bring it closer to the industry's normal pace. One of the people said the company plans to release an updated version of its V4 model, V4.1, in June.
DeepSeek’s new model V4.1 will provide enterprise users with more tools and better support for the MCP protocol. A person familiar with the matter said that DeepSeek also plans to allow V4.1 to directly process image and audio information, although the content it generates is still limited to text.
DeepSeek’s current round of financing was launched in April, and its expected valuation has risen sharply from the early stage of over US$10 billion (approximately RMB 68 billion). On April 22, The Information reported, citing four people familiar with the matter, that Tencent and Alibaba were negotiating to invest in DeepSeek.
After the completion of this round of financing, DeepSeek’s valuation may exceed 350 billion yuan.
Despite the rapidly soaring valuation, some investors who have been in talks with DeepSeek have expressed concerns about DeepSeek's ability to generate sustainable revenue and accelerate model release. DeepSeek has always been known as a research laboratory, prioritizing model development over commercialization.
DeepSeek is working to turn technology into products and services that can be billed, and its employees have been promoting its model to businesses across a wide range of industries, hoping to spark interest in collaboration. DeepSeek is also recruiting from ByteDance’s product development team and other domestic technology companies, according to two people familiar with the matter.
The Information believes that this round of financing and commercialization measures marks a major strategic shift for DeepSeek. Liang Wenfeng is known in the industry as a technical idealist. He had hoped to keep DeepSeek away from commercial pressure. Many in China's AI community attribute DeepSeek's success to the lack of outside investors and the resulting lack of pressure to commercialize the technology. However, as the cost of computing power soars and the competition for top talent intensifies, this model is becoming increasingly unsustainable. The recent loss of many researchers at DeepSeek is an example.
By bringing in external funds, DeepSeek can invest more in computing power and provide higher salaries to its employees. DeepSeek may also start issuing stock options in the future as a supplement to employee compensation.