DeepSeek's Funding Round Doubles to $12 Billion as Tencent and CATL Anchor It — and Term Sheets Could Push It to 100 Billion Yuan

DeepSeek is close to securing at least 80 billion yuan (about $12 billion) in a fresh funding round, Bloomberg reported on Monday, citing people familiar with the matter — a sum roughly double the 50 billion yuan the Chinese AI lab is said to have targeted when the round began. The financing has not yet closed, and signed term sheets could bring the final total close to 100 billion yuan.

Two of China's largest companies anchor the raise. Tencent, which operates WeChat, and Contemporary Amperex Technology (CATL), the world's biggest EV battery maker, have committed among the largest amounts in the round, according to Bloomberg's sources. Neither figure was broken out. The report also leaves the final valuation unset and does not identify a lead investor — unusual for a round of this size, and a sign of how founder-controlled the process remains.

That control is by design. Founder and chief executive Liang Wenfeng built DeepSeek's research operation on the profits and computing resources of High-Flyer, the quantitative investment firm he co-founded, after studying engineering at Zhejiang University and applying machine learning to trading before founding DeepSeek in 2023. When the lab took its first major external money, Reuters reported in June that investors placed their capital into a limited partnership managed by Liang rather than directly into the company — an arrangement widely read as a way for the founder to keep decision-making power while still absorbing outside capital. Bloomberg's reporting does not say whether the new round uses the same structure.

Why the target moved: DeepSeek's own September 10 announcement introduced V4.1-Flash, a model the company described as faster and cheaper to serve, with a new architecture and lower API prices. Bloomberg's sources say investor interest exceeded expectations after the release, though the report does not connect the model to any specific revenue. What is known on the commercial side is a milestone that circulated in late September: Liang told investors the lab had reached a $1 billion annualized revenue run-rate, according to The Information. A run rate extrapolates a recent revenue pace; it is not a full-year result.

The funding math is striking either way. DeepSeek's first external round was reported at about 50 billion yuan, with Tencent and CATL already among the largest outside investors. Only months later, the target for the next financing is at least 80 billion yuan. Investors are increasing exposure as DeepSeek keeps releasing models and building commercial revenue, while Liang has to balance that money against the independence that defined the lab's early years.

There is also a public-market endpoint in sight. Bloomberg's sources describe a possible landmark IPO in early 2027 — timing that would follow this round by a year or so. Earlier reporting, which aireport.net covered in September, pointed to a Shanghai listing valued at roughly $7.5 billion; the new report does not specify an exchange, and the sources caution that timing could still shift.

What it means: the round would put one of China's most watched AI labs on a capital footing that starts to resemble the frontier labs in the United States, where nine- and ten-figure raises have become routine. But the more telling detail is structural — a founder who routes outside money through a vehicle he manages, a round with no named lead, and a valuation left unstated. The open question for the next few weeks is whether the term sheets close at 80 billion yuan, close to 100 billion, or somewhere in between.