Technology

China's DeepSeek Nears a $15 Billion Deal, Even as the U.S. Pulls Its Tools

Martin HollowayPublished 4w ago3 min readBased on 7 sources
Reading level
China's DeepSeek Nears a $15 Billion Deal, Even as the U.S. Pulls Its Tools
source:deepseek.com

DeepSeek, a Chinese AI company, is close to finishing a fundraising round worth about $15 billion, or roughly 100 billion yuan. Reuters reported on October 22, 2026 that the round was nearing its close. Bloomberg News first reported the raise, citing sources.

TechCrunch reported that the round came two days after the U.S. told federal civilian agencies to remove DeepSeek's AI technology from their systems over security concerns. Potential investors include CATL and Tencent, among others.

The round is the latest in a series of reported funding steps over the past five months. In May, China's newly founded national AI fund was leading investment talks. In June, DeepSeek raised more than $7 billion at a $50 billion valuation. A valuation is the price investors place on the whole company.

In July, the company was reportedly seeking another $1.5 billion at a valuation of around $71 billion. In early October, CNBC reported that DeepSeek was considering expanding its latest round to as much as 100 billion yuan, twice its earlier target. The same day, Reuters reported that the company was set to raise more than 80 billion yuan, or $11.93 billion. The 100 billion yuan figure now matches the October 22 round.

DeepSeek was also reported in July to be planning a possible IPO, which is a company's first sale of shares to the public, in China as early as next year. That plan has not been completed. The current round is a private funding round, so shares go to selected investors, not the public.

DeepSeek was founded in 2023 by Liang Wenfeng, who holds an estimated 81.6% stake in the company's Chinese operating entity after its earlier funding round. The copyright holder named on DeepSeek's website is Hangzhou DeepSeek Artificial Intelligence Basic Technology Research Co., Ltd.

DeepSeek has also updated its DeepSeek V4.1 Flash model to version V4.1-Flash-1017. The company says the update brings much better project development at lower cost and faster speed. It has added API support for web search, and for web scraping and fetching, which means collecting content from web pages automatically. An API is like a service window in a restaurant kitchen: other software places an order, and the model sends back an answer.

The broader context here is two signals arriving in the same week. The U.S. government is removing DeepSeek models from its systems, while investors in China are putting money in at a scale that would have been hard to imagine for a lab founded only in 2023. For companies that build tools on DeepSeek, the practical question is less about the price tag and more about what they must change.

In my view, the thing to watch is whether the new web search and fetching features, along with the cost and speed claims, attract developers outside government. The funding will decide how much computing power DeepSeek can buy. The updates suggest where that power will go: AI programs that work with live web information, offered through an API that developers can call directly.