DeepSeek is in talks to raise its first round of venture capital, with its potential valuation soaring from $20 billion to $45 billion in just a few weeks, the Financial Times and Bloomberg reported.
Chinese AI labs rose to prominence in early 2025 by launching large-scale language models trained at a fraction of the computational power and a fraction of the cost of leading U.S. models like OpenAI and Anthropic. Since then, it has maintained a moderate pace with the world’s top models in areas such as reasoning and coding while maintaining open weight (a version is available for free on Hugging Face).
The institute, founded by Chinese hedge fund billionaire Liang Wenfeng, who controls nearly 90% of the company, had not previously sought investors, the FT reported. But in the face of competitors poaching DeepSeek’s researchers, Liang opted to raise money to offer employees shares in the company, people told the FT.
The round is said to be led by state investment agency China Integrated Circuit Industry Investment Fund, according to a report by Bloomberg. China is seeking to finance homegrown AI technology to circumvent difficulties in obtaining American technology, especially chips. DeepSeek is optimized to run on chips made by Chinese hardware giant Huawei. The duo is seen as a powerful duo for the nation to develop its own AI to rival that of the United States. According to Bloomberg, China’s cloud giants Tencent and Alibaba are also reportedly in talks to participate.
DeepSeek could not be reached for comment.
