bloomberg reported that the startup paused its deal and was exploring an in-house inference chip program, with potential IPO discussions.
DeepSeek told prospective investors it is suspending the deal for now. The Hangzhou-based AI startup had been seeking a valuation of about 500 billion yuan, following a maiden round that raised about $7.4 billion.
Bloomberg News reported the pause and said it followed founder Liang Wenfeng's frustration over online reports of comments he made during the first fundraising round.
The pause comes as DeepSeek pursues an in-house inference chip program and had started early deliberations on a potential IPO on Shanghai's STAR Market, moves that increase the capital demands behind production-scale inference.
DeepSeek rose to prominence with its large R1 reasoning model and an open-weight approach, and over the last year the company has shifted from avoiding outside capital to pursuing sizable external funding to support model development and hardware ambitions.
The company has been recruiting chip‑design engineers and engaging external design houses and suppliers as part of a multi‑quarter push to reduce reliance on third‑party accelerators and cut per‑inference costs, a strategy that lengthens technical timelines and raises supply‑chain risk.
Bloomberg said DeepSeek might resume the deal process at a later date.
