By Hamza Faisal

DeepSeek just raised one of the largest amounts of private capital in Chinese history. The Hangzhou-based AI lab has closed a fresh funding round of about 100 billion yuan — roughly $15 billion — with Tencent and battery giant CATL among its biggest backers, Bloomberg reported on October 22.

The round makes DeepSeek one of the best-funded AI startups on the planet and signals that China's most famous AI lab intends to fight the frontier AI race with a war chest, not just clever engineering.

The target kept getting bigger

DeepSeek didn't set out to raise this much. According to people familiar with the matter cited by Bloomberg, the company originally aimed for 50 billion yuan, then lifted the target to 80 billion yuan as investor demand swelled, before settling at a final tally of 100 billion yuan.

That swell is striking given how young the company's fundraising history is. Combined with an earlier 50-billion-yuan raise, DeepSeek has now secured about 150 billion yuan — over $22 billion — in roughly half a year, according to the report. For a company that only recently began taking serious outside capital, the pace is extraordinary.

The biggest checks came from two very different corners of Chinese industry: Contemporary Amperex Technology (CATL), the world's largest EV battery maker, and Tencent, the WeChat operator with a sprawling gaming, cloud and AI empire. Their presence locks in strategic exposure to what Beijing increasingly treats as a national AI champion.

Still chasing the American frontier labs

Even $15 billion does not close the gap with Silicon Valley. Bloomberg noted that Anthropic and OpenAI have each raised more than $100 billion — many times what DeepSeek has secured. But the imbalance is beginning to moderate, and DeepSeek has one advantage money can't fully buy: efficiency.

The company triggered what markets called the "DeepSeek moment" in 2025, when its models matched much of the performance of leading American systems at a fraction of the training cost — a breakthrough linked to about $1 trillion in wiped-out global equity value as investors repriced the cost of frontier AI. Its V4.1 Flash release in September reportedly pressured listed Chinese AI peers like MiniMax and Zhipu AI.

Reuters, which also carried the Bloomberg report, noted it could not immediately verify the figures — but the story of DeepSeek's ambitions does not depend on the final number. The company is reportedly eyeing a listing, which would be one of the most anticipated Chinese market debuts in years, and rival Moonshot is preparing its own IPO.

What the money buys

Compute is the obvious answer. Frontier model training now consumes data-center capacity on a scale that once belonged to nation-states, and DeepSeek's strategy of cheaper, more efficient models still requires vast GPU fleets to serve hundreds of millions of users. A 100-billion-yuan war chest funds both sides of that bet: more models, and more machines to run them.

The raise also lands as the global AI race splits into two tracks — American labs burning tens of billions of dollars per quarter, and Chinese labs trying to match them with smaller budgets and tighter engineering. DeepSeek's pitch has always been that talent and algorithms matter more than raw spend. With $15 billion in fresh capital, it no longer has to choose between them.

By Hamza Faisal