DeepSeek Founder Liang Wenfeng’s Fortune Soars Into Global Top 50
Liang Wenfeng founded DeepSeek in 2023 after developing artificial-intelligence capabilities within his quantitative hedge fund, High-Flyer. The Hangzhou-based startup gained international prominence with large language models that delivered performance comparable with leading US systems at lower reported costs. Because Liang retained an unusually large stake while financing the company largely with trading profits, DeepSeek’s rising private-market valuation has translated directly into personal wealth and become a marker of investor confidence in China’s AI sector.
Forbes estimated Liang’s net worth at $39.5 billion as of Aug. 12, 2026, up about 3,850% from roughly $1 billion a year earlier and placing him among the world’s 50 richest people. The jump followed DeepSeek’s first external funding round in June, when it raised about 50 billion yuan ($7.4 billion) at a post-money valuation of roughly 400 billion yuan. Liang contributed about 20 billion yuan himself, helping him preserve control despite the arrival of state and corporate investors.
All Coverage
1 original reportsThe Backstory
The history behind this eventDeepSeek Raises More Than $6.7 Billion at Valuation Above $50 Billion
DeepSeek, founded by Liang Wenfeng, who has a quantitative-fund background, has quickly emerged as a focal point of China’s AI industry with its low-cost large language models. Amid U.S. chip export controls and China’s push for homegrown AI technology, a valuation above $50 billion would not only make it China’s most valuable AI startup but also reflect investors’ high expectations for domestic foundation models.
As of July 20, 2026, DeepSeek had raised more than $6.7 billion in its first funding round from investors including Tencent, CATL and Chinese state-backed funds. Liang retained control of the company through a special partnership structure. After completing the round, the company immediately announced plans to invest 50 billion yuan in expanding its team, with a focus on developing AI agents and next-generation training models.
Subscribe to Mark Radar Weekly
Every Friday, the week's strongest signals in your inbox. Unsubscribe anytime.