Kimi’s capital boom is dragging China’s AI into a second arms race

The dark side of the Chinese artificial intelligence company is considering first listing in Hong Kong and then landing in Shanghai. Reuters cited news on September 10 that the star start-up behind Kimi is exploring the double listing path in Hong Kong, and previously reports suggested it could raise about $3 billion in Hong Kong. China-generated AI competition from model performance competition, rapidly upgraded to a comprehensive competition for computing power, finance, talent and capital market seats.

Unlike traditional Internet companies, leading-edge big model have to consume huge computing costs until their business models are fully mature. The biggest enemy of AI startups may not be no users, but the model means more expensive chips, training clusters, and talent costs. Therefore, an IPO for the dark side of the month is not merely a reward for wealth, but more like continuing to keep fuel on the table.

From Kimi to the capital market, speed itself is a signal.

The dark side of the moon relies on Kimi’s long text capabilities to quickly enter the public eye, and the new generation Kimi K3 is trying to expand its influence in open weight and intelligence capabilities. At the same time, DeepSeek, Ali, bytes, intelligence spectrum, MiniMax and other companies are competing for model entry. China’s AI industry has formed a cruel logic: model leadership can only last for months, and the financing window decides whether the enterprise can withstand the next round of iteration.

The Hong Kong market can connect international funds and global investors, while the research board is linked to China’s policy capital, technology industry chain and domestic valuation system. The financing path of an AI company is being designed as a two-set system of “international capital + national science and technology strategy” to supply blood simultaneously.

As model competition reaches billions or even tens of billions of dollars, so-called “startups” are becoming more and more like an infrastructure project that needs capital markets, industrial policies and national computing to be supported jointly.

Behind the IPO boom, there’s another ignored question: Who pays for failure?

Capital markets pursue AI, which can accelerate innovation and potentially create valuation bubbles. Large-model enterprises are currently facing problems that do not match commercial revenues and training costs. Once listed in the industry joint exports, investors ultimately buy not only technical imagination, but also high computing power procurement, continued financing and the risk of model rapid depreciation.

China's regulators have recently asked traders not to push a large number of low-quality companies to the IPO market, which forms a subtle contrast to the listing of AI companies. Policy is clearly willing to provide a wider capital channel for "strategic technology", but strategic labels can not replace profitability and governance transparency. The dark side of the month in the future if listed, the market will really need to pay attention not only to the model list, but to pay users, reasoning costs, revenue structure, associated transactions and calculation dependence.

China’s AI capital race has at least four observation points

• Can Hong Kong become a major financing window for Chinese AI companies to bypass geopolitical pressures;
• Will TechCrunch continue to give big model companies higher policy valuation;
• Can the decline in computing costs catch up with model revenue growth;
• Will U.S. chip restrictions force listed companies to disclose higher supply chain risks?

On the one hand, China hopes that AI companies will grow rapidly and catch up with the United States in global model competition; on the other hand, over-financing and homogeneous competition may replicate the new energy car industry’s price war.

Focus China believes that the Kimi Sprint Capital Market marks China’s AI entry into the second phase: the first phase is competing for modeling capabilities, and the second phase is competing for who will get enough capital to survive the next round. The real winners will not only be the companies that can finance the most, but also companies that can prove that technology, income and governance can withstand the test of the open market.