China’s AI firms struggle with funding as US rivals pull ahead
Photo: Tima Miroshnichenko / Pexels
Anthropic's record IPO highlights a widening capital gap that Chinese AI startups are finding hard to close
The US-China race for artificial intelligence dominance has a new bottleneck, and it isn’t talent or ideas. It’s money.
Chinese AI companies are finding it harder to raise the capital needed to train increasingly powerful models.
The numbers behind the gap
Between 2023 and 2026, US AI firms secured about $380 billion in venture funding. Chinese startups pulled in approximately $39 billion over the same stretch.
Anthropic’s IPO, which took place in November 2026, raised over $81 billion and gave the company a valuation of $1.5 trillion.
China has not been entirely shut out of big deals. DeepSeek raised approximately $15 billion (100 billion yuan) in October 2026, one of the largest AI funding rounds in the country.
Public markets aren’t helping
For Chinese AI firms that have already gone public, investors have been unkind. Zhipu AI, which also operates under the Z.AI brand, has seen its shares fall as much as 85%. MiniMax shares are down about 34%.
Falling share prices matter beyond bruised portfolios. A sinking stock makes it more expensive to raise fresh capital, because every new share sold brings in less money.
AI, tech, and the markets they move—in one daily briefing.
Daily. Free. Join 34,000+ readers across crypto, finance, and policy.
Tencent reported third-quarter 2026 revenue of 207.3 billion yuan, an 8% increase from a year earlier. That growth still fell short of expectations, and operating profit dropped 7%.
Alibaba and Baidu are also among the major Chinese firms contending with the same pressures as they compete against well-capitalized American labs.
Chips, not just cash
Training frontier AI models requires vast numbers of advanced chips, and Nvidia’s are the industry’s workhorse. US export controls have already limited Chinese access to the most capable processors.
In November 2026, Washington went further. New curbs restrict Chinese companies from remotely accessing advanced Nvidia chips, closing off a workaround that had allowed some firms to rent computing power hosted elsewhere.
The impact falls unevenly. Startups are being hit harder than the big hyperscalers, which have more existing infrastructure, deeper balance sheets and more ways to absorb the shock.
What this means
For investors, the picture points to a riskier environment around Chinese AI stocks. The sharp declines at Zhipu AI and MiniMax suggest markets are already pricing in a tougher road ahead, and continued regulatory pressure from Washington could add to that volatility.
Tencent’s results show the pressure isn’t limited to startups. When a company of Tencent’s size sees operating profit slip while pouring money into AI, it signals that the spending race is straining even the biggest balance sheets. Watch upcoming earnings from Alibaba and Baidu for similar strain on margins and cash flow.
There are a few signals worth tracking from here. One is whether DeepSeek’s October round proves to be an outlier or the start of larger domestic financing. Another is how Chinese firms adapt to the November chip restrictions, and whether they can sustain model training without remote access to advanced Nvidia hardware.
Disclosure: This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our Editorial Policy.

