DeepSeek pauses second funding round after leaked founder comments on Nvidia reliance and US AI gap
DeepSeek has paused its second funding round, targeting at least $1.4 billion at a $71-74 billion valuation, after unverified transcripts of founder Liang Wenfeng's investor remarks spread on Chinese social media. Liang reportedly said DeepSeek still relies on Nvidia $NVDA chips.
DeepSeek has suspended its second funding round, verbally telling prospective investors on July 25 that it would not sign agreements in the coming days. The round had targeted at least 10 billion yuan, or about $1.4 billion, at a pre-money valuation between 480 billion and 500 billion yuan ($71-74 billion), following a record $7 billion first round in June. The pause followed unverified transcripts of founder Liang Wenfeng's private meetings with investors that circulated on Chinese social media and technology outlets. In the leaked remarks, Liang reportedly said China's AI industry still trails the United States in funding and computing infrastructure rather than talent, and acknowledged DeepSeek remains reliant on Nvidia $NVDA hardware, unable to train its largest models at the scale of leading US developers. DeepSeek is separately preparing an IPO on Shanghai's STAR Market that could still happen this year despite the funding pause. For a company whose reputation rests on efficient, lower-cost AI development, an admission of continued Nvidia dependency lands at a sensitive moment ahead of a planned public listing.