Editorial illustration for DeepSeek Seeks More Capital Weeks After USD 7B Funding Round
DeepSeek Seeks New Funding After $7B Round
DeepSeek closed a $7 billion funding round in late May at a $52 billion valuation. Barely two months later, the Hangzhou-based AI lab is back in front of investors, this time talking about a pre-money valuation of roughly $71 billion. The Financial Times reports the company is in early discussions with new backers to fund a build-out of its own data centers and secure AI chips, a departure from the cloud-rental model most labs lean on.
Founder Liang Wenfeng put up about $3 billion of the first round himself, making him DeepSeek's largest single investor. CATL, Tencent, JD.com, NetEase, and a state-backed AI fund rounded out the rest. The new money would go toward physical infrastructure and toward an inference chip DeepSeek is developing in-house, aimed at cutting its dependence on Nvidia and Huawei hardware.
The urgency traces back to how DeepSeek has been competing on price. Its V4 models, including V4-Pro and V4-Flash, have been priced low enough to pull in a wave of US business customers, but that strategy burns cash fast and requires the kind of compute capacity DeepSeek doesn't yet own outright.
DeepSeek is already raising again. The Chinese AI lab just closed its first funding round and needs capital for its own data centers and chips to keep its aggressive pricing strategy going.
Why this matters
For developers and founders building on cheap inference, DeepSeek's math is worth watching closely. A company burning through $7 billion in weeks, then going back to investors for a round at ten times that valuation, tells us the V4 pricing strategy isn't sustainable on current margins. CATL, Tencent, JD.com, and NetEase aren't charities, they're betting that subsidized tokens buy market share DeepSeek can monetize later, the way Uber and Amazon Web Services once did.
If you've built products assuming DeepSeek's rock-bottom API prices are permanent, they might not be. That's a real dependency risk, not just a pricing footnote. For researchers, the more interesting signal is competitive: Zhipu and other domestic rivals are closing the gap fast enough that DeepSeek needs data centers and chips it doesn't yet own, rather than just better models.
Capital intensity is creeping into a market that sold itself on efficiency gains over brute-force compute. Watch whether this round actually closes at $71 billion, and whether V4-Pro's 1.6 trillion parameters translate into performance gains that justify the spend, or just bigger bills for everyone downstream.
Related Reading
Business & Startups Kling launches Video O1, all-in-one model with MVL bridge using transformer
Business & Startups Enterprise voice AI splits into three architectures, shaping compliance
Business & Startups Nordic pilot adds Gemini for Education, NotebookLM to boost AI literacy
From Archives Anthropic's New AI Ad Campaign Draws Criticism for 'Creepy' Tactics
From Archives Reflection Signs USD 1 Billion Compute Deal With Nebius
Common Questions Answered
Why is DeepSeek seeking additional capital just weeks after closing its $7 billion funding round?
DeepSeek is raising more funds to build out its own data centers and secure AI chips, moving away from the cloud-rental model used by most AI labs. This infrastructure investment is necessary to sustain the company's aggressive pricing strategy for its V4 models, which the article suggests is not currently profitable on existing margins.
What is DeepSeek's valuation in the new funding round compared to its May valuation?
DeepSeek's pre-money valuation in the new funding discussions is approximately $71 billion, up significantly from the $52 billion valuation it achieved in its May funding round just two months prior. This represents a substantial increase in company valuation in a very short timeframe.
How much capital did founder Liang Wenfeng contribute to DeepSeek's first funding round?
Founder Liang Wenfeng personally invested approximately $3 billion into DeepSeek's initial $7 billion funding round in late May. This substantial personal investment demonstrates the founder's significant commitment to the company's growth and development.
What is the business model strategy behind DeepSeek's subsidized token pricing?
DeepSeek's strategy of offering heavily subsidized tokens is designed to acquire market share quickly, similar to how companies like Uber and Amazon Web Services built their user bases through aggressive pricing. The company is betting that it can monetize this large user base later once it has established market dominance and reduced competition.
Which major Chinese companies are backing DeepSeek's infrastructure and pricing strategy?
CATL, Tencent, JD.com, and NetEase are among the major backers supporting DeepSeek's subsidized pricing model and infrastructure development. These companies are investing in DeepSeek's strategy as a calculated bet that the market share gains will eventually become profitable and valuable.
Further Reading
- DeepSeek Just Raised $7.4 Billion. Here's The Catch. - Forbes
- DeepSeek's $10.29B Financing Round Explained - DeepInfra
- DeepSeek nears first funding round as investors scramble for ... - Yangtzeer
- DeepSeek slated to raise $7 billion in maiden funding round, sources say - Reuters
- DeepSeek could hit $45B valuation from its first investment round - TechCrunch
