DeepSeek reportedly raises $15 billion days after US federal ban

What's going on

Chinese AI lab DeepSeek has raised around 100 billion yuan, roughly $15 billion, according to a Bloomberg report. That would make it one of the largest funding rounds in China's history. Bloomberg cites people familiar with the matter. DeepSeek has not yet confirmed the round, and The Next Web stresses that it has not independently verified the report. TechCrunch asked DeepSeek for comment but did not immediately hear back.

Investors and how the round unfolded

The report names battery maker CATL and Tencent, the internet giant behind WeChat, as possible backers, along with others. According to Bloomberg, investor demand pushed the round higher during talks: DeepSeek initially aimed for 50 billion yuan, raised its target to 80 billion and finally closed at about 100 billion.

It is the company's second outside round this year. In June, DeepSeek raised more than $7 billion at a $50 billion valuation. In July, it reportedly sought another $1.5 billion at a valuation of around $71 billion. Earlier reports also said an IPO in China may be planned as soon as next year.

The political backdrop

The round came two days after an order from the US Department of Homeland Security (DHS): civilian federal agencies must stop buying DeepSeek products and remove them from their systems. The stated reasons are security concerns, including the risk that Chinese authorities could gain access to US government data. One of the sources (TechBuzz) reads the timing as an act of defiance and cites alleged ties to military research institutions; according to that source, the company denies such allegations. The reports offer no evidence of a causal link between the ban and the funding, especially since rounds like this take a long time to negotiate.

What the money is for

The Hangzhou-based company is reportedly stepping up spending on AI infrastructure and open models (open-weight models, whose trained weights anyone can download and adapt) for domestic and international markets. Its cloud service is said to already run on chips from Chinese maker Huawei, despite US chip export restrictions. The expansion could intensify competition among Chinese rivals even further.

Comparison and background

DeepSeek was founded in 2023 by Liang Wenfeng, who also co-founded the hedge fund High-Flyer, which funded the lab's early development. The lab became known worldwide in early 2025 with R1, a free open-weight reasoning model that showed strong models can be built at a fraction of the usual cost, and that sent tech stocks tumbling. According to WireScreen, Liang holds an estimated 81.6% of the Chinese operating company after the earlier round.

Our take

The sum shows that investors are betting heavily on Chinese AI labs despite US restrictions. Still, DeepSeek remains far smaller financially than OpenAI and Anthropic, which have each raised more than $100 billion, according to Bloomberg. It is unclear who exactly is investing, at what valuation, and how much of the money has actually been paid in. The strongly opinionated passages about "national AI champions," state funds or talent poaching come from TechBuzz and are speculation, not established fact. For users of open models, the main takeaway is that DeepSeek is likely to have more funds for data centers and new open models.