Last year, DeepSeek surprised everyone by building a competitive AI model on a fraction of the usual budget. They were lean, smart, and tiny – about 200 people.
Not anymore.
In mid‑June, they closed their first external funding round: $7.4 billion, at a valuation over $50 billion. That’s the biggest single AI funding round in China so far.
Founder Liang Wenfeng personally put in ~$3 billion. Tencent, CATL, NetEase, JD.com, and IDG Capital joined. But Liang kept full control – outside investors get no voting rights and can't exit for five years.
What's the first move? Hiring – everywhere.
DeepSeek just posted 33 roles across 7 departments: engineering, algorithms, product, even HR and legal. They want to at least double every team. Offices in Beijing, Hangzhou, and Inner Mongolia. Interns too.
Their pitch: "We give newbies core tasks from day one. And we don't look for geniuses – we look for people who shine in their own way." One job posting already got 10,000+ resumes by Friday afternoon.
From scrappy to scaled
This isn't just growth – it's a pivot. DeepSeek used to brag about doing more with less. Now they're spending big to catch up in headcount. Compare: ByteDance's Doubao has 2,000 people, Anthropic 3,000, OpenAI 5,000+. Even after doubling, DeepSeek will still be smaller – but the direction is clear.
They're also getting serious about making money. New product and client‑facing roles, plus a dedicated "Harness" team for building AI agents, signal a shift from pure research to commercial play.
What could go wrong?
Fast scaling kills culture. Decision‑making slows, communication gets messy. Can they keep their edge?
Chip sanctions are still biting – they're working with Huawei on domestic alternatives, but that's years behind Nvidia.
And monetisation? No Chinese AI firm has cracked that yet. Investors are already pushing for revenue.
Bottom line
DeepSeek built its name on efficiency. Now they're spending big to grow. Whether that gamble pays off – we'll know in 12–24 months.