According to Forbes, DeepSeek founder Liang Wenfeng has debuted on China’s 100 richest list with an $11.5 billion fortune after his company rattled the AI world in January 2025 with its low-cost R1 model. The 40-year-old’s wealth comes largely from his stake in the privately held firm, which Forbes values at $15 billion. DeepSeek hasn’t tapped external investors and is mostly funded by Liang’s quantitative hedge fund High-Flyer, which manages about $8 billion in assets. The company’s estimated sales are only a fraction of OpenAI’s reported $12 billion in annualized revenue, despite DeepSeek’s app having over 73 million active users in September. Meanwhile, analysts value DeepSeek between 5% and 10% of OpenAI’s $300 billion March 2025 valuation.
Sponsored content — provided for informational and promotional purposes.
The cost advantage
Here’s what makes DeepSeek so fascinating: they’re spending pennies where others are spending fortunes. The company published research showing they trained one AI model for just $294,000. Compare that to the reported $500 million OpenAI spent on a single GPT-5 training run. That’s basically building a skyscraper for the price of a shed. But there’s a catch – that $294,000 figure doesn’t include the estimated $1.6 billion DeepSeek has spent on hardware over the years, including Nvidia chips they can no longer easily access due to U.S. restrictions.
The commercialization challenge
Now here’s the problem DeepSeek faces: making money in China’s AI market is tough. While OpenAI charges $20-200 monthly for ChatGPT and makes serious revenue from developers, Chinese companies mostly offer their chatbots for free. DeepSeek charges business developers just $0.03 per million tokens compared to OpenAI’s $0.125. That’s a huge price difference, but when you’re competing in a market where free is the expectation, even low prices might not be enough. And with U.S. chip restrictions forcing them to use Huawei hardware, their cost structure is getting more complicated.
The competition heats up
So can DeepSeek maintain its momentum? The competition is getting fierce. Alibaba is planning to invest over $50 billion in AI over the next three years, and analysts note that developers are starting to lean toward Alibaba’s Qwen models, which some say are “just as good as DeepSeek’s models, if not better.” When you’re up against that kind of financial firepower, being the low-cost innovator only gets you so far. The company also hasn’t released a next-generation model yet, just incremental updates to existing ones.
What’s next for DeepSeek?
Liang has been pretty clear that his primary motive is driving innovation rather than hitting financial targets. He told local media that “China has been a follower for too long” and wants to push technological boundaries. But innovation doesn’t pay the bills forever. The world is waiting to see if DeepSeek can translate its technical achievements into sustainable business success. Can they maintain their cost advantage while scaling up? Will they eventually need to take outside funding? These are the billion-dollar questions facing China’s newest AI billionaire.
