DeepSeek's $15 Billion Fundraise Doubles Its Target as China Chases the US

The Biggest AI Fundraise of the Year

DeepSeek is wrapping up a funding round of about 100 billion yuan, roughly $15 billion, according to Bloomberg reporting published October 22. That is double the 50 billion yuan the Chinese AI startup originally targeted, and it would make this one of the largest private funding rounds in AI history.

The investor list tells its own story. The biggest backers are battery maker CATL and WeChat operator Tencent, according to people familiar with the matter. New investors include home appliance maker Midea, food delivery giant Meituan, automaker Geely, and semiconductor maker GigaDevice. Boyu Capital, CPE, and Legend Capital are in, alongside existing shareholder IDG Capital. More than 15 investors are participating in total.

The DeepSeek fundraise values the company at 500 billion yuan before new capital, about $70 billion, per Reuters. The round is expected to close in the coming days, though the sources cautioned the details are not yet ironclad.

Why Industrial China Is Buying Into AI

Look at who is writing the checks and a pattern emerges. CATL makes batteries. Midea makes appliances. Geely makes cars. GigaDevice makes chips. These are not software VCs betting on the next app; they are industrial giants buying exposure to the AI models they expect to run their factories, vehicles, and devices. This is the physical economy placing its wager on intelligence as infrastructure.

That framing matters because it is different from how the US AI race is funded. American labs raise from venture firms and hyperscalers. DeepSeek is raising from the companies that will deploy its models into hardware. Several of the investors are backed by the state, which fits Beijing's stated aim of getting ahead of the US in AI. When your battery supplier, your appliance maker, and your automaker all want a piece of your AI lab, the lab stops being a software startup and starts looking like national infrastructure.

The IPO Clock Is Ticking

The DeepSeek fundraise also signals that the company is getting closer to public markets. In September, Reuters reported that DeepSeek had hired CITIC Securities to prepare for a potential STAR Market IPO, Shanghai's tech-focused board. A $15 billion private round at a $70 billion valuation is exactly the kind of pre-IPO positioning that sets up a blockbuster listing.

The timing is aggressive but legible. DeepSeek went from research curiosity to global name in under two years, and its open-weight releases have repeatedly embarrassed larger, better-funded Western labs on efficiency. The open-weight release of DeepSeek V4.1 Flash earlier this month kept that reputation intact. Money is now chasing the story, and the story is that China can do frontier AI at a fraction of Silicon Valley's cost.

What the DeepSeek Fundraise Buys

Compute, mostly. Training frontier models is a capital expenditure business now, and $15 billion buys a great deal of GPUs, data center capacity, and engineering talent. It also buys time: with this war chest, DeepSeek does not need to monetize aggressively or rush an IPO on anyone else's schedule.

The round also sharpens a contrast with Western AI funding. While US labs raise enormous rounds too, the legal AI boom and physical-world AI bets like ASI show capital fragmenting across verticals. DeepSeek's raise is horizontal: one lab, one bet, the biggest checkbook in Chinese tech.

There are real risks buried in the celebration. A $70 billion valuation demands frontier results indefinitely, and export controls still constrain the chips Chinese labs can buy. Geopolitics is not a footnote here; it is the business model. But the investor lineup suggests China's industrial base has already decided which horse it is riding. When the companies that make the world's batteries, cars, and appliances all pile into the same AI lab, the debate about whether AI matters to the physical economy is over. The only remaining question is who captures the value.