Racing toward an IPO at a $50 billion valuation, why does Kimi dare to take its story to the secondary market?

Kimi is reportedly preparing a confidential Hong Kong listing application, but what really deserves attention is not “when it will go public,” but what logic it plans to use to win secondary-market pricing.
It is reported that Moonshot AI secretly submitted an A1 application to the Hong Kong Stock Exchange this week, while advancing a new funding round at a pre-money valuation of approximately $50 billion. The company has not officially confirmed the information.
Why is this figure sensitive?
Because AI company valuations are undergoing an extremely rapid repricing. Moonshot AI’s valuation had previously continued to rise, reportedly reaching approximately $35 billion post-money after completing financing in July. Now the market is again citing a pre-money valuation of $50 billion.
In other words, capital markets are no longer valuing AI companies based solely on “whether they have a model,” but on whether that model can create a commercial flywheel.
Where does Kimi have an advantage?
First is brand recognition. Kimi initially opened up the market with its long-context capabilities, building a certain user base among ordinary users, students, developers, and knowledge workers.
Second is the speed of model iteration. The launch of K3 has refocused market attention on Moonshot AI’s technological capabilities and commercial potential.
Third is revenue growth. Previous public reports showed that Kimi’s ARR grew rapidly in the first half of 2026, indicating at least that AI products are not limited to user scale and are beginning to show clearer paid conversion.
But the problems are equally clear.
A $50 billion valuation is not a reward, but a report card.
After the listing, investors will not permanently grant you an AI premium simply because your name is Kimi. The market will ultimately break down your revenue growth, gross margin, R&D expenses, computing costs, and cash flow.
In particular, the most common misconception in the large-model industry is that “rapid user growth = a mature business model.”
In fact, the two are completely different.
If computing costs also rise as the number of users increases, scale may be a double-edged sword; only when revenue growth begins to outpace cost growth will an AI company truly have a sustainable business model.
So my greatest expectation for Kimi’s IPO is not whether it can create yet another Hong Kong-listed star stock, but whether it can tell the market:
When exactly can China’s large models move from an “valuation story” to a “cash flow story”?
If Kimi can provide an answer, the significance of this IPO will extend far beyond one company’s listing.#Kimi秘密交表启动港股IPO