Two days. That's how long it took DeepSeek to answer Washington with money. The US government had just ordered federal civilian agencies to strip the Chinese startup's AI tools from their systems over national security concerns. Then DeepSeek closed a $15 billion funding round — one of the largest private AI raises in Chinese history.
The cash didn't come from Silicon Valley. It came from Chinese industrial heavyweights. CATL, the world's biggest battery maker, doubled down with another $1.38 billion after putting in $687.5 million earlier. Tencent joined in, alongside chipmaker GigaDevice, food-delivery giant Meituan, carmaker Geely, appliance maker Midea and internet group NetEase. More than fifteen investors signed on in total, including venture firms Boyu Capital, CPE, Legend Capital and IDG Capital.
DeepSeek's valuation climbed to roughly 500 billion yuan, around $70 billion — twice what the company was aiming for just months ago, when it first raised $7.4 billion from outside investors at a target valuation near $74 billion. This time, backers put real money behind that number.
For the company that rattled chipmaker stocks in early 2025 by proving a cheap model could compete with the best, the round carries a double message. It shows Chinese corporations are willing to bankroll homegrown AI even as Washington cuts off access to federal contracts. And it shows DeepSeek's open models still pull serious money months into a brutal race against OpenAI, Google and Anthropic.
Omdia analyst Lian Jye Su called the valuation a vote of confidence in DeepSeek's open-source strategy. The company has already hired CITIC Securities to prepare for a listing on Shanghai's STAR Market, with an IPO possibly landing before year-end. The new money is earmarked for model research and compute — the very resource Washington is trying to keep out of its reach.
