<a href="https://www.google.com/preferences/source?q=finance.biggo.com" rel="nofollow noopener" target=”_blank”>Add preferred source
Chinese AI startup DeepSeek has locked in at least RMB 80 billion (approximately $12.0 billion) in its latest funding round, with the final total potentially reaching RMB 100 billion (approximately $14.9 billion)—far exceeding the original RMB 50 billion target. Tencent Holdings and CATL are the lead investors, with the capital earmarked for AI business expansion and IPO preparation in early 2027. DeepSeek plans to deploy at least 160,000 Huawei high-end AI accelerators in Inner Mongolia and has open-sourced infrastructure components supporting Huawei’s Ascend platform, accelerating China’s AI ecosystem shift away from Nvidia dependence. Following the round’s completion, the company will undergo a corporate restructuring to meet STAR Market listing requirements.
Key Elements
Chinese artificial intelligence startup DeepSeek’s latest funding round has dramatically exceeded its original target, locking in at least RMB 80 billion (approximately $12.0 billion), with the final total potentially reaching RMB 100 billion (approximately $14.9 billion). Chinese battery giant CATL and internet behemoth Tencent Holdings (0700.HK) are the lead investors, and the capital will pave the way for DeepSeek’s initial public offering (IPO) planned for early 2027.
DeepSeek initially set a fundraising target of approximately RMB 50 billion (approximately $7.5 billion), but after releasing its latest V4.1-Flash model in September, market response was overwhelmingly positive and investor interest far exceeded expectations. According to people familiar with the matter, based on signed investment term sheets, the final financing total could approach RMB 100 billion. The fundraising process is now nearing completion, though details may still be adjusted during final negotiations.
Two Industrial Giants Place Simultaneous Bets
The lead investor lineup in this round reflects how China’s major tech platforms and industrial capital are simultaneously doubling down on the AI foundation model space. Tencent had already invested RMB 10 billion (approximately $1.5 billion) in the previous round, making it the largest single investor at the time, with the goal of integrating AI into its social and content platforms. CATL (300750.SZ), the Chinese battery leader, is likely eyeing the growing component supplier opportunities in the AI supply chain, including data center hardware demand.
DeepSeek completed its first external funding round in June this year, raising RMB 51 billion (approximately $7.6 billion) at a post-money valuation of nearly RMB 400 billion (approximately $59.8 billion), setting a record for the largest single-round financing in China’s AI industry. That round was led by founder Liang Wenfeng, with Tencent, CATL, NetEase, JD.com, IDG Capital, and other industrial and financial investors participating, alongside a strategic follow-on investment from a national-level AI industry investment fund. Liang Wenfeng’s personal net worth doubled after the first round, reaching approximately $36 billion.
From Model Competition to Capital Warfare
Analysts note that DeepSeek, which previously built its reputation on a “low-cost, high-efficiency” approach, has now entered a phase of hyperscale capital expenditure—signaling that China’s AI competition has escalated from a pure model capability race to a full-scale capital war encompassing computing power, data centers, chips, and talent.
DeepSeek plans to build a large data center in Inner Mongolia, deploying at least 160,000 high-end AI accelerators from Huawei Technologies. If executed as planned, this would become one of the largest known Huawei AI chip clusters to date. DeepSeek has also officially open-sourced infrastructure components for Huawei’s Ascend computing platform, extending the partnership from hardware procurement into joint software ecosystem development.
Beyond meeting its own rapidly growing computing needs, this initiative is viewed as a landmark case for China’s AI industry advancing a domestic computing infrastructure. DeepSeek CEO Liang Wenfeng has told investors that increasing the proportion of Chinese-made chips used in model training is a key objective, and he expects Huawei to begin supplying training-grade chips as early as Q4 2026.
Taking Aim at Nvidia’s Software Ecosystem Moat
DeepSeek’s deepening collaboration with Huawei directly challenges Nvidia’s most enduring competitive advantage in the Chinese AI market: its software ecosystem. Nvidia’s global dominance stems not only from hardware performance but also from its mature CUDA software environment, which gives developers compelling reasons to remain on the Nvidia platform.
DeepSeek has re-engineered its TileLang tool to support Huawei’s Ascend platform and released it as openims to lower the switching costs for Chinese AI companies migrating from Nvidia platforms to domestic computing infrastructure, enabling smoother transitions across models, software, and hardware
However, Chinese tech analysts point out that while the software gap is narrowing, the chip supply shortfall has not improved in tandem. Huawei has acknowledged difficulty meeting the rapidly growing demands of China’s AI industry, while Chinese tech giants such as Alibaba and ByteDance are reportedly still queuing to purchase Nvidia chips. The existence of a substantial gray market and smuggling networks also indicates that actual demand for Nvidia products in the Chinese market remains robust.
IPO Preparation and Outlook
DeepSeek has engaged CITIC Securities (600030.SS) as its IPO advisory institution, preparing for a listing on the Shanghai Stock Exchange’s STAR Market. The specific listing timeline, corporate valuation, and fundraising size have yet to be finalized, but market expectations broadly point to early 2027.
Following the completion of this funding round, DeepSeek’s next step will be a corporate restructuring to comply with listing requirements. Rival Moonshot AI is likewise targeting an early 2027 IPO, and after completing a financing round at a $50 billion valuation, the capital race among Chinese AI startups is heating up across the board.
DeepSeek has rapidly built its reputation in the global AI race since launching a product in 2025 that achieved capabilities comparable to Silicon Valley models at a fraction of the cost. The newly released V4 Flash has further redefined market expectations for AI model price-performance, directly challenging U.S. AI companies such as Anthropic and OpenAI. Liang Wenfeng has made clear in meetings with investors that the company will continue developing open AI models while pursuing the long-term goal of artificial general intelligence (AGI), rather than prioritizing commercial monetization.
Once added, BigGo Finance appears first in Google Search Top Stories, so you get the broadest, most up-to-the-minute, and most comprehensive global financial news first.
