From Kimi to Hong Kong stocks, AI investment is moving from “storytelling” into the “age of doing the math”

If the most popular term in the AI industry over the past two years was “model capabilities,” then the market may care about only three words next: Can it make money?
Kimi’s parent company, Moonshot AI, has reportedly secretly submitted an A1 filing to the Hong Kong Stock Exchange while advancing a new funding round at a pre-money valuation of approximately $50 billion. Although the company has not confirmed the IPO news, this move alone is enough to make the market reassess the pace at which Chinese AI companies are going public
Because the large-model industry has entered its second phase.
The first phase was about the models.
Whoever had larger parameters, stronger reasoning, and longer context windows was more likely to attract attention.
The second phase is about commercialization.
Are users willing to pay? Are companies willing to purchase? Can API revenue grow? Can model usage translate into stable cash flow?
These are the questions listed companies must answer.
Previous market reports indicated that Kimi’s ARR grew rapidly in the first half of 2026, surpassing $100 million in March and $300 million by mid-June, a remarkably rapid pace of growth.
If this growth trend can continue, then the IPO would indeed have a relatively strong foundation for its story.
But I want to point out one thing: Revenue growth for AI companies does not equal profit growth.
This is precisely what makes large models unique—the marginal cost of adding one more user to a traditional internet product may be very low; but every additional complex inference by AI may correspond to costs for computing power, electricity, chips, and cloud services.
So going forward, when capital markets evaluate Kimi, they cannot look only at ARR.
They must also look at a more brutal metric: how much must be burned to earn every dollar.
If revenue grows rapidly while unit costs continue to decline, the business model will look increasingly attractive; conversely, if revenue growth is driven by greater investment in computing power, the valuation must be approached with greater caution.
That is also why Kimi’s IPO is worth watching.
It could become a window into China’s AI industry moving from “technology valuations” to “commercial valuations.”
The market was previously willing to pay for the future; now it is beginning to demand that the future gradually appear in financial statements.
So I believe the biggest highlight of Kimi’s listing is not whether it will become Hong Kong stocks’ new “AI star,” but whether it can answer the question facing the entire industry:
Are large models ultimately an arms race involving unlimited cash burn, or a business that can eventually make money sustainably?
If Kimi can prove the latter, the truly valuable aspect of this IPO may not be the stock of a listed company, but the valuation paradigm it establishes for China’s entire AI industry.
#Kimi秘密交表启动港股IPO