DeepSeek, the Chinese AI lab known for its open source models, is closing in on the largest funding round in its history. According to reports citing Bloomberg on October 6, 2026, the company is set to raise at least 80 billion yuan, about $12 billion, in a round led by Tencent and battery giant CATL. The raise comes ahead of a planned 2027 listing that is expected to be one of China's biggest stock debuts in years.
DeepSeek Expands Its Round From 50 to 80 Billion Yuan
The round initially targeted about 50 billion yuan, but investor demand grew after the company released its latest models, pushing the size to at least 80 billion yuan with a chance of approaching 100 billion yuan. These figures are not yet an official announcement, but they show how willing Chinese capital is to back the country's leading AI players.
Tencent and CATL Anchor DeepSeek's Shareholder Base
In an earlier round this year, DeepSeek raised 50 billion yuan at a valuation of roughly $52 billion to $59 billion. Founder Liang Wenfeng contributed 20 billion yuan himself, Tencent put in about 10 billion yuan and CATL about 5 billion yuan, alongside JD.com, NetEase and China's national AI investment fund. Reports say Liang's net worth has doubled to around $36 billion on the back of the higher valuation.
Fresh Capital Targets 160,000 Huawei Chips
A large share of the money is headed for infrastructure. DeepSeek plans to use about 160,000 Huawei AI chips in a new data center in Inner Mongolia and is writing custom code to run its models efficiently on that hardware. On the listing, the company has reportedly hired CITIC Securities for a potential debut on Shanghai's STAR Market, though timing, valuation and deal size have not been set.
What It Means for Investors: Appetite for Chinese AI Remains Strong
A round that has grown more than 60% beyond its original target signals that investors see DeepSeek as a flagship of Chinese AI. For most investors, the closest exposure today is through backers such as Tencent and CATL. Keep in mind that every figure still comes from sources rather than the company, and that building out large amounts of compute is expensive in a fiercely competitive Chinese AI market.
