DeepSeek wants $74 bn cash before onshore IPO
Chinese AI startup DeepSeek is preparing to initiate a new fundraising round, targeting a valuation of approximately 500 billion yuan ahead of a possible initial public offering on the mainland, according to sources. The plan emerges shortly after the Hangzhou-based company, which garnered international recognition for its affordable AI models in 2025, secured approximately $7.4 billion in June at a post-money valuation of around 450 billion yuan, according to sources. Filings by two Chinese investors later suggested DeepSeek was valued at 350.88 billion yuan, or around $52 billion. The consecutive fundraising initiatives highlight a robust investor interest in one of China’s most scrutinised AI firms, while simultaneously indicating the escalating expenses associated with AI competition, which necessitates substantial computing power, data-center infrastructure, and engineering expertise.
DeepSeek is seeking to secure up to 50 billion yuan in its latest funding round, as reported by a source. Early deliberations have commenced regarding a potential IPO on Shanghai’s Nasdaq-style STAR Market, according to sources. The company has established an internal objective to finalise an IPO filing within this year, according to one source. All individuals chose to remain anonymous due to the non-public nature of the information. The fundraising and IPO plans are currently in preliminary stages, and the terms and timetable are subject to change, they indicated. DeepSeek did not provide an immediate response to a request for comment. Source reported that DeepSeek was preparing for a potential IPO filing, while the another source indicated that the company was considering a new fundraising round at a valuation of no less than 480 billion yuan.
DeepSeek disrupted global technology markets last year with the introduction of models that seemed to compete with top US systems while offering reduced training and operating expenses. Soon after its maiden fundraising round in June, DeepSeek announced plans to double its workforce across various departments, including areas such as data centers and AI agents, which are systems capable of performing tasks with limited prompting. Several of those initiatives will necessitate substantial capital outlay. Earlier this month, source reported that DeepSeek was seeking to develop its own AI inference chip and had quietly ramped up the hiring of chip-design engineers for the initiative. DeepSeek had long distinguished itself in China’s AI sector by eschewing external funding. Founder Liang Wenfeng had primarily financed the company through his quantitative hedge fund High-Flyer prior to its recent external funding, sources previously informed.
However, the expense associated with maintaining a leading position in AI has increased significantly, necessitating a strategic shift. DeepSeek has encountered significant competition over the past year from domestic tech behemoths like ByteDance and Alibaba, alongside well-capitalized AI startups such as Z.ai, Moonshot, and MiniMax. In the June funding round, DeepSeek founder Liang personally committed 20 billion yuan, while Tencent Holdings and battery giant CATL contributed 10 billion yuan and 5 billion yuan respectively, thus becoming the largest external shareholders, as reported by source. Additional investors comprise China’s national AI fund, gaming developer NetEase, and e-commerce giant JD.com, alongside investment firms IDG Capital, Loyal Valley Capital, Monolith Management, and Shixiang Capital, as reported by various sources. The involvement of the state-backed AI fund underscored DeepSeek’s strategic significance to Beijing’s initiatives aimed at fostering domestic AI leaders and diminishing dependence on foreign technology.
