Liang Wenfeng

Liang Wenfeng (梁文锋; born 1985) is a Chinese entrepreneur and businessman. He co-founded the quantitative hedge fund High-Flyer and is the founder and chief executive of the artificial intelligence company DeepSeek, which is based in Hangzhou, China. His career moved from applying machine learning to finance to building large-scale AI research, and he became widely known outside China after DeepSeek released its models in 2025.

Early life and education

Sources differ on Liang's birthplace: some name Wuchuan county in Guangdong province, and others name Zhanjiang City, of which Wuchuan is a part. Both of his parents were reportedly primary school teachers. He entered Zhejiang University in 2002 at about the age of seventeen to study electronic information engineering, and he earned a bachelor's degree followed by a master's degree in information and communication engineering. His master's research concerned target-tracking algorithms that used low-cost pan-tilt-zoom cameras, a topic within computer vision. Sources give different years for his degrees, and the accounts of his early career are largely based on interviews and press profiles.

Founding High-Flyer

After university, Liang initially pursued experiments in applying AI to various fields, which were not commercially successful. He turned to finance, and in 2013 he and his Zhejiang classmate Xu Jin set up an investment management firm in Hangzhou. In 2015 or 2016 (sources differ), he and two classmates founded Ningbo High-Flyer Quantitative Investment Management Partnership, which is generally known as High-Flyer. The firm used mathematical models and machine learning to make trading decisions in financial markets. Liang has described a quantitative fund as one in which investment decisions are made by models rather than by people. By 2019 the firm reportedly managed more than 10 billion yuan, and by 2021 its assets under management were reported to exceed 100 billion yuan. In 2019 Liang gave a keynote address at the Golden Bull Awards on the future of quantitative investment in China.

Shift toward artificial intelligence

High-Flyer's growing computing needs led the firm to build its own computing infrastructure. Around 2021 Liang reportedly began purchasing thousands of GPUs from Nvidia, a purchase that some of his business partners regarded as ambitious. Regulatory pressure on the Chinese financial sector between 2019 and 2023, as reported by press accounts, encouraged the team to treat AI research as a side project. Liang wrote the preface to the Chinese edition of a 2021 book about the quantitative investor Jim Simons, whose approach he has cited as an influence.

DeepSeek

Liang announced in 2023 that High-Flyer would devote resources to AI research, and DeepSeek was established that year as a separate company. Sources differ on the founding month, with some giving July 2023 and others May 2023. DeepSeek's stated goal, according to a 2023 interview with the Chinese technology publication 36Kr, is artificial general intelligence, meaning AI that matches or exceeds human cognitive ability across tasks. The company has been funded largely by High-Flyer rather than by venture capital, and reports suggest that venture firms showed little interest in it at the start.

DeepSeek released the open-source mixture-of-experts model DeepSeek-V3 in December 2024, which was reported to perform comparably to GPT-4o on several evaluations. Its large language model DeepSeek-R1 followed in January 2025, and its release drew wide attention, including a notable sell-off in technology stocks in the United States. DeepSeek's V2 model, released in mid-2024, is reported to have started a price war among Chinese AI providers. Liang has argued that a developing Chinese economy should contribute to open technology rather than benefit from others' work, and that closed-source models could delay competition only temporarily.

Public profile and recognition

Liang has maintained a low public profile and has rarely given interviews. Media accounts describe him as an engineer-turned-investor who led a small, largely domestically trained team. Foreign and Chinese outlets have reported on his role as a symbol of Chinese AI competitiveness, and his hometown has attracted visitors. In July 2026 Bloomberg reportedly ranked him as the wealthiest founder of an AI model company, with an estimated net worth of about 36 billion US dollars, although such valuations are based on estimates and vary across sources.

Debates and uncertainties

Discussion of Liang and DeepSeek has focused on the reported low cost of training its models, which some analysts questioned, and on the role of Chinese state policy and US export controls on advanced chips in shaping the company's computing resources. Figures for High-Flyer's assets, DeepSeek's founding date, and Liang's personal history come mainly from press reports, company statements, and secondary profiles, and they are not always consistent across sources.