DeepSeek raises ~$7.4B at $50B+ valuation with investor no-poaching clause

DeepSeek closed its first external funding round at roughly $7.4 billion (50 billion yuan), valuing the Hangzhou lab at more than $50 billion — by several reports as high as $59 billion — and making it China's most valuable AI-exclusive startup. Founder and CEO Liang Wenfeng personally committed 20 billion yuan (about $3 billion) to the round, an unusually large insider stake that signals his intent to retain control.

The most striking term is a no-poaching clause: per CNBC, DeepSeek told investors they may not recruit its employees. In a market where AI talent commands soaring premiums — and where DeepSeek's lean, research-heavy team is its core asset — the clause is a defensive moat as much as a financing condition. Backers include Tencent, battery giant CATL, and the state-linked National AI Industry Investment Fund, underscoring both private and strategic state interest.

The raise follows DeepSeek's release of DeepSeek-V4-Pro, a 1.6-trillion-parameter mixture-of-experts model, and lands as the company navigates U.S. scrutiny: separate reporting indicated Washington is holding off — for now — on blacklisting DeepSeek among more than 100 firms deemed potential security risks. That regulatory overhang is the key caveat for Western enterprises eyeing DeepSeek models.

The deal feeds this week's louder theme — China's open and frontier labs accelerating on domestic silicon. With GLM-5.2 topping open-weights leaderboards and DeepSeek now flush with capital, the gap-narrowing narrative that began with the original DeepSeek-R1 moment is back in force. Watch whether the valuation holds up against U.S. export-control and blacklist decisions still pending.