DeepSeek, the Hangzhou research lab that upended assumptions about the cost of frontier AI, is close to sealing one of the largest private funding rounds in Chinese technology. Reports this week put the raise at roughly 50 billion yuan, about $7.4 billion, at a pre-money valuation near $74 billion.
The financing was expected to close by the end of August, according to China Money Network and Tech Startups. A full raise at that price would leave the company worth roughly $81 billion once the money lands, placing it among the most valuable AI firms outside the United States.
Who is writing the cheques
Existing backers are leading the round. They include the investment firm Monolith, Shixiang Capital, and the battery giant Contemporary Amperex Technology, better known as CATL, per Capacity. CATL's presence is a reminder of how closely China's AI ambitions are tied to its industrial base, where power, manufacturing and compute increasingly sit under one roof.
Much of the new capital is earmarked for compute. DeepSeek built its reputation on training strong models at a fraction of the budgets its American rivals spend, but the next generation of systems still needs graphics chips, data centres and electricity at scale. The raise reads as an admission that efficiency alone does not make the hardware bill disappear.
An IPO on the horizon
People familiar with the plans say DeepSeek could file for an initial public offering as soon as the end of 2026, with a listing on Shanghai's STAR Market slated for 2027. That would make it one of the first of China's new AI champions to test public markets at home, a route that keeps the company clear of the regulatory friction a US listing would invite.
The timing matters. DeepSeek's earlier models pressured pricing across the industry and helped push open-weight releases from rivals such as Z.ai. A well-funded DeepSeek preparing for public markets suggests the competition is about to get sharper, not gentler.
Commentarii · 0