Chinese AI lab DeepSeek has raised about $15 billion - roughly 100 billion yuan - in one of the country's largest-ever funding rounds, Bloomberg reported on October 22. Potential investors named in the report include CATL, the Chinese battery maker, and Tencent, the company behind WeChat, among others - reported as potential backers, not confirmed participants.
The timing is striking. Bloomberg's report was published just two days after the United States ordered federal civilian agencies to remove DeepSeek's AI technology from their systems, citing security concerns. The report does not say when the round closed, so the raise may predate the restriction. The ban applies to federal civilian systems - not the broader US market - but it walls DeepSeek out of federal civilian use and signals to American enterprises that adopting its models carries political risk. DeepSeek did not immediately respond to TechCrunch's request for comment on the report.
Whether or not the round closed after the restriction, the juxtaposition is stark: Washington is pushing DeepSeek out of federal civilian systems while Chinese capital funds it at a scale few American labs have reached. For the reported potential investors, the bet appears to be that the domestic Chinese market - plus international markets where DeepSeek's open-weight models can circulate freely - is big enough to justify a nine-figure-yuan check.
If confirmed, the round would be the latest step in a fast-moving financing story. DeepSeek raised more than $7 billion in June at a $50 billion valuation, and was reportedly seeking an additional $1.5 billion in July at a valuation around $71 billion, with plans for a potential IPO in China as early as next year. Founder Liang Wenfeng retains an estimated 81.6% stake in the company's Chinese operating entity after the earlier round, according to corporate ownership data analyzed by WireScreen and cited by TechCrunch.
DeepSeek, based in Hangzhou, was founded in 2023 by Liang, who previously co-founded the quantitative hedge fund High-Flyer - which helped finance the lab's early development. That origin matters: unlike most AI labs, DeepSeek began life funded by trading profits rather than venture capital, which let it operate for its first two years without answering to outside investors. The company gained global attention in early 2025 when it released its free-to-use, open-weight R1 reasoning model, a release that challenged assumptions about how much capital and compute frontier AI required and put pressure on US labs' dominance. R1's efficiency claims triggered a broad repricing of AI infrastructure stocks at the time and made DeepSeek a household name well beyond China.
According to the report, DeepSeek is ramping up spending on AI infrastructure and open models for both domestic and international markets, even as it faces restrictions in the US. The new capital would fund that expansion and could intensify competition among China's domestic AI labs. Export controls limit the advanced chips Chinese labs can buy, so infrastructure spending on this scale also raises a practical question the report does not answer: what hardware that money can actually purchase. Whatever the answer, the open-model strategy is central. By releasing weights anyone can download, DeepSeek builds global adoption that restrictions on its US commercial presence cannot easily reach - developers worldwide can run its models without DeepSeek ever selling them anything.
The raise fits a broader pattern of enormous checks flowing into AI companies worldwide - and into Chinese labs in particular. Moonshot AI, the company behind Kimi, secured $2 billion at a $20 billion valuation in May. Manus AI raised more than $500 million earlier this month. Zhipu AI (Z.ai) and MiniMax, which both went public in Hong Kong in January, have market capitalizations of roughly $29 billion and $7.2 billion respectively as of today. A $15 billion DeepSeek round would dwarf all of these and reset the benchmark for private AI financing in China.
For the global AI race, a $15 billion DeepSeek round would signal that US export controls and agency bans are not slowing Chinese frontier-lab funding - if anything, domestic capital and strategic investors like CATL and Tencent are stepping in at a scale that rivals the largest American rounds. For startups building on open models, it also suggests DeepSeek's open-weight releases will keep coming, backed by a war chest large enough to sustain expensive training runs for years. That is good news for builders who want capable models without per-token API fees, and uncomfortable news for US labs whose pricing power depends on open models staying a generation behind the frontier.
Caveats apply: the round is a Bloomberg report, not a company announcement, and DeepSeek has not confirmed it. The final size, valuation and investor list could change when - and if - the company makes the round official. The reported valuation for this round has not been disclosed; the $71 billion figure circulating in July was attached to a smaller, separate raise. What is clear is the direction: the world's best-known open-weight lab is being funded to compete at frontier scale.