The most visible example of this strategy is CXMT Corp., China's largest maker of DRAM memory chips. In July 2026, the company raised 57.92 billion yuan ($8.6 billion) in Asia's biggest IPO of the year, selling 6.69 billion shares at 8.66 yuan each on Shanghai's tech-heavy STAR Market . When trading began, shares surged roughly 466% — the largest first-day jump among the world's 10 biggest IPOs by deal value in 2026 . The stock closed at 49 yuan, lifting CXMT's market capitalization to about 3.3 trillion yuan ($488 billion), instantly making it the most valuable listed company in mainland China . This was all powered by public-market investor demand, not direct government handouts .
The company, which held a 7.67% share of the global DRAM market in 2025, plans to use the IPO proceeds mainly for mass-producing memory wafers . Its DRAM chips power everything from AI data centers to smartphones, positioning CXMT as a critical link in China's AI hardware supply chain .
Perhaps the most talked-about AI startup in China is DeepSeek. In July 2026, reports emerged that the Hangzhou-based AI pioneer is planning a fresh fundraising round at a valuation of about 500 billion yuan ($74 billion) ahead of a potential initial public offering on Shanghai's STAR Market . The company raised $7.4 billion in June 2026 alone in its first-ever outside funding round, backed by investors including Tencent, battery giant CATL, and Beijing's National Artificial Intelligence Industry Investment Fund .
Multiple news reports indicate DeepSeek has begun formal IPO preparations, engaging accounting and banking advisors, with a possible filing by late 2026 or early 2027 and a public debut in 2027 . For a company that just months earlier was valued at around $50 billion, this trajectory underscores how private and public capital — not government grants — are now funding frontier AI research in China .
The capital-market-driven model extends beyond CXMT and DeepSeek. Robotics maker Unitree is preparing a STAR Market listing with DeepSeek as a key strategic investor, taking about 934,000 shares worth 141 million yuan (roughly $20 million) in the offering . This cross-pollination of the so-called "Hangzhou Six Tigers" startup ecosystem showcases how public markets are becoming the primary exit and scaling vehicle for China's next-generation technology companies .
China's AI funding shift is amplified by a structural cost-of-capital advantage over the US. While US hyperscalers plan $370 billion in AI capital expenditure this year, China's top tech firms — Tencent, Baidu, Alibaba, and ByteDance — will invest less than 10% of that amount . From 2023 to 2025, combined capex by Chinese hyperscalers totaled $124 billion, roughly 82% less than the $694 billion spent by US peers .
Yet this leaner approach is yielding competitive results. By some estimates, China's top AI models now achieve about 90% of US performance levels, despite far lower spending . Chinese companies benefit from lower domestic labor costs, subsidized industrial power, and a pragmatic strategy that favors open-source models and broad adoption over capital-intensive frontier spending . Moonshot AI's Kimi K2 cost just $4.6 million to train — a fraction of the billions that OpenAI spends on R&D . This efficiency-focused approach allows Chinese firms to achieve competitive AI outcomes at a fraction of US capital outlay .
This is not simply a funding shift — it represents a fundamental change in how the Chinese state approaches industrial policy. By mobilizing its $28 trillion stock and bond markets, Beijing is creating a self-sustaining ecosystem where successful AI and semiconductor companies can tap public demand for capital, rather than relying on direct government spending . The CXMT IPO alone demonstrated the immense appetite among mainland Chinese investors for AI-related stocks, and the pipeline of startups waiting to follow suit suggests this model has only just begun .
As DeepSeek, Unitree, and other "Hangzhou Six Tigers" startups prepare their public listings, one pattern is clear: China's AI funding model has entered a new, capital-market-driven phase — one that is leaner, faster, and increasingly independent of traditional state subsidies.