I snagged fake Kimi legacy shares.

As Kimi's financing pace accelerates, the trading of Kimi's existing shares has become extremely hot.

At the beginning of July, an "explosive news" popped up on FA Y's WeChat: Kimi is transferring shares at an external valuation of 180 billion US dollars, and is raising the next round of financing at a valuation of 315 billion US dollars. Regarding the source of the existing shares, the information he got is that Yang Zhilin himself, together with his childhood friend, personally set up a fund. The terms are: a 10% management fee charged upfront, a 20% carry drawn at the back end, and "no third parties are allowed to participate".

Despite the harsh terms, the news is attractive enough. Out of trust in the information source, Y did not verify it too much and immediately took action: confirming the GP list, connecting funds, and preparing for subscription. There are too many institutions wanting to participate, and some of the list includes well-known institutions and industrial parties.

But the story line broke at the last minute. The funds were all ready, but the agreements and TS could not be delivered for a long time. "The matching between funds and shares is not working out," Y said dejectedly, and the news could not be verified.

I also heard the news of "existing share transfer from Yang Zhilin" from another source. The other party said: "The acquirer has a strong background, there will be no more than 3 intermediaries in the process, the quota and price change in real time, the success rate is 80%, and it is very sought-after."

Another intermediary revealed that the quota "increases in price at any time". He said that the round at 180 billion US dollars was gone because the decision-maker hesitated for a while, and then the new round of price was received. "Although we don't know if the news is true, the basic quota is all bought out as soon as it comes out, and it is impossible to find it outside."

Fast pace, uncertainty of price and quota, and rampant false information can almost summarize the core characteristics of this wave of Kimi existing share transactions. The most exaggerated false news includes ordinary retail investors claiming that they have bought Kimi's pre-listing shares; even in the same group, there is a phenomenon where people expose each other's news as false.

The situation is so fierce that Kimi has to personally issue a clarification announcement: all its financing activities are solely the direct responsibility of the company, and emphasizes that "the transfer of the company's existing shares (including common shares, incentive equity, etc.) must be approved by the company internally before being carried out. All existing share transactions that occur without the company's approval are invalid."

For people engaged in existing share transactions, valuation is only one side of the story. What they really care about is liquidity and certainty. When an existing share transaction cannot even provide a TS, but there is still a long queue of buyers, it shows that the hunger of the capital side has overwhelmed the instinct of due diligence.

$500 Billion Financing: Pre-IPO Round Opens Even Before Series F Closes

If the recent rumors about DeepSeek's share transfer in the market can be 100% judged as false news, the news of Kimi's share transfer is half-believed and half-true in the market.

People believe that Liang Wenfeng of DeepSeek holds more than 80% of the shares himself. In addition, the shares of external investors in DeepSeek's first round of financing are locked for 5 years and cannot be sold, so "existing share circulation" is basically impossible, and the shortage of shares is determined by its equity structure. However, Kimi has more shareholders, more financing rounds, and more scattered chips, so the tradable volume of existing shares is significantly higher. Decentralization means that any "mysterious existing share" is more difficult to be falsified, which is not only the reason for the frantic rush for quotas, but also easier to breed "quota illusion".

Recently, the news of Kimi's new shares and existing shares quotas spreading everywhere in the market has reached its peak. The situation can be glimpsed from the financing pace in recent months.

From January to February 2026, Kimi completed three consecutive rounds of financing, with scales of 5 billion, 7 billion and 7 billion US dollars respectively, and the valuation was pushed from 100 billion US dollars to 180 billion US dollars. In May, Meituan Longzhu led the investment of about 20 billion US dollars in Series D, and the post-investment valuation exceeded 200 billion US dollars. In June, a new round of financing was launched, and the pre-investment valuation jumped to 315 billion US dollars.

The real explosion point was in July. Two weeks after the release of Kimi K3, on July 29, Series F was completed, with a financing amount of over 35 billion US dollars, and the post-investment valuation was locked at 350 billion US dollars. The original fundraising target of this round was only 10 billion to 20 billion US dollars, and the final capital entering the market was more than 3 times over, so the round was closed early directly.

The over-subscription of Series F has also accelerated the subsequent pace a lot. The Series G (Pre-IPO), which was originally scheduled to start in August, was launched ahead of schedule, and the pre-investment valuation was directly set at 500 billion US dollars (about 3.5 trillion RMB). Goldman Sachs and CICC have served as joint sponsors, and the company will go public in Hong Kong in the fastest 6 months.

Another source said that the latest round of financing also has clear requirements for institutions: participants need to complete the payment before August 15; if you want to be directly included in the company's captable, the investor needs to be an institution with a management scale of more than 5 billion US dollars.

From 43 billion US dollars of Series C at the end of last year to 500 billion US dollars of Pre-IPO, the valuation has increased by more than 10 times in less than a year. There is also news that the fastest Hong Kong IPO will be launched in 6 months. Although the official denied the news of listing application, IPO is only a matter of sooner or later.

An insider who participated in Kimi's financing mentioned that the rapid rise of Kimi's valuation not only benefited from the market sensation of the K3 product, but also came from the significant acceleration of the commercialization process. In the second quarter of 2026, ARR rose continuously from 100 million US dollars in March to 300 million US dollars in June, doubling month-on-month; API revenue accounted for more than 70%, overseas market revenue increased by 400% year-on-year, and the business covered more than 200 countries. In addition, within 20 days after the release of K2.5, its revenue exceeded the total amount of the whole year of 2025, and the price increase strategy did not affect the demand, so the commercialization has entered a stage of rapid volume growth.

Undoubtedly, Kimi's shareholder lineup is absolutely top-tier. HSG has participated in every round of financing since the angel round. IDG Capital led the investment of 150 million US dollars in Series C. 5Y Capital led the investment of more than 7 billion US dollars in February 2026. ZhenFund, Capital Today, and L.S. Capital were present from the very first day. Gaorong Ventures entered from Series B, and Source Code Capital increased its stake in Series D. KAIYU Capital, China Merchants China Fund (via Minxing Investment), and CPE Yuanfeng all have shares in it. It is almost impossible to see a project gather so many top US dollar VCs outside the AGI track.

There are also industrial capital giants making strategic layouts. Alibaba led the Series A+ round of Kimi, and Meituan Longzhu, Tencent, and Xiaohongshu entered the market one after another. Among them, Alibaba and Tencent participated in almost every round, Meituan Longzhu led the investment from Series A to Series D, and the individual strategic investor Wang Huiwen has invested a total of nearly 100 million US dollars. These are not financial investors who come to earn multiples, but have the intention of synergy with e-commerce, local life, content, and social scenarios. Kimi's C-end traffic is highly complementary to these super app ecosystems, and holding shares means occupying a strategic position.

State-owned background funds and international long-term capital will naturally not be absent. The Shenzhen Hexie Growth Phase III Fund participated by the National Social Security Fund holds 11.76% of Kimi's shares; in addition, China Mobile participated in the investment through its fund, and the National Artificial Intelligence Investment and Management Co., Ltd. and Beijing Artificial Intelligence Industry Investment Fund entered the market one after another. This also conveys the narrative of "national strategic asset".

A person from a domestic top institution that participated in Anthropic's financing last year said bluntly, among domestic model companies, if you close your eyes and choose two to invest directly, "one is DeepSeek, and the other is Kimi". The core reason is that one of the common points of these two companies is their overseas reputation, which is exactly the core variable to attract long-term capital.

Another front-line institutional person V also revealed to me that they entered the market at a valuation of about 60 billion US dollars, and continued to follow up the next two rounds of investment. The reason why they continued to follow up is: "The model is good, the product is good, the cognition is good, the globalization is well done, and the team responds quickly."

In March this year, the overseas product Cursor claimed that its product was independently developed, but was caught "wrapping" Kimi, and finally the official confirmed that it was an authorized commercial cooperation. This incident is a strong proof of Kimi's globalization capability.

Chase, or Not to Chase?

Under the heat, the mentality of old shareholders is far more complicated than that of new capital.

Take the institution where V works as an example. The luck factor for them to enter the market is no less than the vision factor. Precisely because they are in the game, their institution did not follow up in this particularly hot Series F and Pre-IPO round, not because they are not optimistic. "We just got a little flustered after calculating the accounts."

Their team believes that the decline in valuation multiples is certain, and the downward pressure on Anthropic's P/ARR is a lesson from the past.

Why will it definitely happen? Because the current valuation is inevitably full of FOMO fanaticism and scarcity premium. But when the company moves towards IPO, information is more transparent, and more players appear, the valuation will return from "storytelling" to "detailed accounting", and the multiple being pressed down is a sure thing. Can Kimi's current $500 billion valuation maintain this crazy multiple after listing? No one can guarantee it.

His words also reveal the collective hidden worry of the primary market in the AI craze: capital is pouring in too fast, and the valuation of each company is more daring to shout; today it is still at a high point, and looking back at Zhipu and MiniMax, their market value has dropped for a round. This track is most afraid of being stuck in the middle. If you don't run forward, the shares will soon be worthless. As for how much you can earn in this game in the future, "I have no idea in my heart."

He also mentioned another factor: "Now they are going to dismantle the red-chip structure."

Dismantling the red-chip structure is the standard action for this round of AI companies to go to the Hong Kong Stock Exchange / A-share market, but few people regard it as a discount factor in valuation. Behind the dismantling of the red-chip structure, it represents a series of invisible frictions to outsiders: how the capital flows, how the shares of US dollar investors return to China in compliance, what the consideration is calculated at, whether the par exit can be implemented, the endless disputes between domestic and foreign lawyers, Kimi's own structure has been revised back and forth for several versions.

Therefore, in the hustle and bustle of over-subscription and flying rumors, they chose to press the pause button.

However, it is certain that after the open source of K3, the confidence of the industry has also increased. With the official open source of the complete model of Kimi K3, the moments on WeChat were swiped instantly. Parameters, rankings, and Musk's praise are all fermenting. Laymen watch the excitement, but in the primary market, people watch the overall situation.

On the day the model was released, an investor M who was doing due diligence on a large model application layer project was in a meeting with the founder of the invested enterprise, and the other party was nervous: "Is this a good thing or a bad thing for us?" His judgment is: it is a good thing.

On the one hand, the open source of the cutting-edge model has taken a big step forward, and the certainty of the infrastructure and application layer is getting higher - the stronger and more open source the model is, the greater the demand for reasoning computing power, and the more certain the "water sellers" such as optical modules, HBM, and data centers will benefit. This is completely consistent with his logic of heavy position in AI infrastructure in the first half of the year.

In addition, Kimi dares to pull the API pricing to the highest level among domestic products, setting a new price anchor for the industry. Previously, everyone was competing for lower prices, and now someone finally proves that "the price is high for a good reason" can also capture the market.

M recalled that three years ago, he missed a very popular star project because of a "wait-and-see attitude". This time, although he didn't catch up, he felt more at ease. It's not that he resigned to fate, but that he figured out that instead of chasing others, it's better to figure out where the next wave of opportunities is. He re-sorted the whole industrial chain from top to bottom. He believes that there are still a lot of cards not shown this year: DeepSeek, MiniMax, and Qianwen are all iterating, and the open source of K3 is just the first shot.

When Kimi's last round of financing is about to close, instead of regretting that the valuation is rising too fast, it is better to see clearly who is the first to cash in on this chain. "For people doing investment, where the cake is and how to cut it is more important than whether you can catch up," M said.

This article is from the WeChat official account "ChinaVenture", author: Li Man, published with authorization from 36Kr.