Six "little dragons" are vying for the title of the first to go public among large model companies.

"Five out of the 'Six Little Dragons' are already preparing for an IPO."

Li Zheng (a pseudonym), a partner at a leading venture capital institution, revealed this news to Zimu Bang (ID: wujicaijing). Following the news that Zhipu is planning an IPO, recent media reports suggest that MiniMax is considering a Hong Kong IPO. According to The Paper, an insider close to MiniMax said that the company does have such an idea, but it is still in the preliminary preparation stage.

Actually, "Five of the 'Six Little Dragons' have long started contacting investment institutions to raise funds, with the scale exceeding $500 million each, and they have all entered the preliminary preparation stage," Li Zheng said.

MiniMax and Zhipu are not the only large - model startups preparing for an IPO. As the domestic listing standards are further relaxed, a situation where five of the 'Six Little Dragons' compete for the title of the 'First Large - Model Stock' is bound to emerge.

According to Reuters, at the 2025 Lujiazui Forum, the China Securities Regulatory Commission (CSRC) announced plans to set up a new board on the STAR Market in Shanghai, which mainly consists of technology stocks, to accommodate unprofitable growth - oriented enterprises. It not only specifically mentioned artificial intelligence companies but also stated that it will restart the application of the fifth set of standards on the STAR Market for unprofitable enterprises. "This means that even large - model startups that are currently still in the red or have insufficient profits have a chance to list on the Growth Enterprise Market," Li Zheng added.

According to a report in Intelligent Emergence, an employee of Baichuan Intelligence remembers that Wang Xiaochuan, the founder of Baichuan, said at an internal meeting last year that the goal for 2025 was to reach the listing threshold (1 billion yuan) in performance, and once on the green - channel list, the company could go public at any time. For the 'Six Little Dragons', it's still uncertain who will be the first to ring the listing bell.

Whether it's the Hong Kong stock market or the A - share market, it is undoubtedly an excellent opportunity for the 'Six Little Dragons' to compete for the title of the 'First Large - Model Stock'. For the latecomers still in the preparation stage, Zhipu, which was the first to spread IPO news, may not necessarily be the final winner.

A

In 2025, the struggling 'Six AI Little Dragons' are all waiting for a chance to "leap over the dragon gate" - to go public.

On April 14, Zhipu officially submitted a listing guidance filing to the Beijing Securities Regulatory Bureau, with China International Capital Corporation as the guidance institution. It became the first company among the 'Six Little Dragons' to initiate the IPO process.

Founded in 2019, one of the key factors for Zhipu being the first to spread IPO news is that it meets the basic requirement of continuous operation for more than three years. MiniMax, founded in December 2021, which was reported to be preparing for an IPO in June, also meets this basic requirement due to its relatively early establishment. However, since the other companies were founded between mid - 2023 and early 2024, they do not meet the requirement of "three - year establishment".

However, Zhipu, which took the lead, may not necessarily become the 'First Large - Model Stock'.

In the past six months, Zimu Bang has learned from investors of different institutions that five of the 'Six Little Dragons' have not clearly advanced their IPO plans but have frequently contacted investors, sending signals of going public. Among them, Zhipu was the first to spread the news of filing.

In terms of seizing time, Zhipu may not have a first - mover advantage. According to Zhipu's filing plan, its listing guidance will be carried out in three stages: completing the preliminary investigation and plan formulation in April; focusing on standardization and rectification and continuous due diligence from May to July; and entering the sprint stage from August to October to prepare listing application documents. That is, after preparing for the IPO and going through necessary procedures such as roadshows, Zhipu will still need another six months to a year.

The other five of the 'Six Little Dragons' are also making similar efforts. After going through these necessary listing procedures, the remaining companies will also meet the basic requirement of three - year establishment.

Meanwhile, there is also a possibility that this requirement will be relaxed.

According to Reuters, at the 2025 Lujiazui Forum, the CSRC announced plans to set up a new board on the STAR Market in Shanghai, which mainly consists of technology stocks, to accommodate unprofitable growth - oriented enterprises. According to the guiding opinions issued on the CSRC's official website, enterprises with the characteristics of "major technological breakthroughs, broad business prospects, and high R & D investment" can list on the new board.

Wu Qing, the chairman of the CSRC, said that the CSRC will set up a new growth layer for science and technology innovation, officially launch the third set of standards for the Growth Enterprise Market, restart the application of the fifth set of standards on the STAR Market for unprofitable enterprises, and support high - quality unprofitable innovative enterprises to list, with a focus on artificial intelligence companies.

Under the relaxed policies, not only do large - model star startups founded for less than three years have hope, but even the long - criticized profitability issue for investors has been pardoned.

For several unprofitable large - model startups, after the restart of the application of the fifth set of standards on the STAR Market for unprofitable enterprises, the pressure of insufficient profits can be temporarily relieved. For the 'Six Little Dragons', this year is undoubtedly the year to decide the 'First Large - Model Stock'.

B

It should be noted that the title of the 'First Large - Model Stock' does not just sound good for the 'Six Little Dragons'.

"The listing logic of the Chinese Internet also applies to AI large models, which is a winner - takes - all situation."

An investor who used to be engaged in FA business told Zimu Bang that both the Internet and AI large - model industries are highly competitive, following the logic of the survival of the fittest. As the first stock in this field, it often enjoys a high valuation premium due to the lack of reference, which can squeeze the living space of other startups.

It is not surprising that the 'Six Little Dragons' are all seeking to go public. This is not only because of the huge capital demand of the startups themselves but also due to the pressure from investors.

"Early - stage investors are waiting to recoup some of their funds, and it's hard to avoid the requirement of a listing time limit in the investment agreements of the 'Six Little Dragons," the above - mentioned investor added.

However, the competition for the 'First Large - Model Stock' is not just about raising funds.

Perhaps SenseTime, Megvii, Yitu, and CloudWalk, which competed for the 'First Stock' during the era of the 'Four AI Little Dragons' in previous years, can answer this question.

In 2021, when AlphaGo sparked public interest in artificial intelligence, four companies led by SenseTime, Megvii, Yitu, and CloudWalk vied for the qualification to be the first to ring the listing bell. Although SenseTime suffered a loss of 24.272 billion yuan in three and a half years, it finally went public at a loss and became the 'First Stock'.

During the IPO, SenseTime's issue price of HK$3.85 was at the lower end of the issue price range. On the first day of listing, SenseTime's stock price rose and then fell back, closing up 7.27%. On the next day, the stock price rose 33.17%, reaching a peak of HK$9.7, and its market value once exceeded HK$320 billion. Among the four companies, SenseTime, the first to go public, was able to keep up with the pace in the large - model era thanks to continuous market financing. In contrast, Megvii shifted to intelligent driving and almost completely abolished its original business; the businesses of companies like CloudWalk have shrunk significantly. For example, CloudWalk laid off a large number of employees in 2024, with R & D personnel accounting for up to 51%.

The 'First Large - Model Stock' has real - world appeal in the market. For the 'Six Little Dragons', the title of the 'First Large - Model Stock' is not only a new story to tell investors but also can actually increase their survival time and space.

Competing for the 'First Large - Model Stock' means being the first among the 'Six Little Dragons' to gain investors' recognition. It is not only easy to obtain a corresponding premium but also can become the price anchor for later - listed companies.

At the same time, for the 'Six Little Dragons' that are in urgent need of capital injection, the most direct benefit of going public is to introduce direct capital support for the enterprise.

In 2025, whether it is the technology layout of tech giants like Tencent, Alibaba, and ByteDance in AI or the technological leadership of DeepSeek, the 'Six Little Dragons' are facing an embarrassing situation of competing with giants and having difficulty telling a good technology story. Moreover, it is becoming increasingly difficult for large - model startups to raise funds from the outside.

Even Zhipu, which still raised funds in the first half of the year, had a single - round financing amount of no more than 1 billion yuan. In March, Zhipu announced the completion of three strategic investments, with amounts exceeding 1 billion yuan, 500 million yuan, and 300 million yuan respectively. According to Tianyancha data, the total disclosed financing amount of Zhipu in 2024 was nearly 7 billion yuan, and in December last year, the single - round financing amount was 3 billion yuan.

For the 'Six Little Dragons' that need to continuously burn money to keep up with technological iterations, it is too difficult to raise funds. So where can the money come from? Perhaps for the 'Six Little Dragons' now, it can only come from the stock market.

C

However, the 'Six Little Dragons' still have a lot to do before going public.

Whether it is the Hong Kong Stock Exchange or the STAR Market, the 'Six Little Dragons' with insufficient profitability urgently need to show their profitability to the outside world.

The STAR Market requires a market - value - centered approach, comprehensively considering factors such as revenue, net profit, R & D investment, and cash flow. Although the main board of the Hong Kong Stock Exchange has introduced a listing system for specialized technology companies, the threshold for non - commercialized companies has been lowered from HK$10 billion to HK$8 billion. However, the requirements for the Growth Enterprise Market (GEM)'s market - value + revenue + cash - flow test are a market value of no less than HK$2 billion at the time of listing, annual revenue of no less than HK$500 million in the past year, and a net cash inflow from operating activities of no less than HK$100 million in the past three years.

The star startups among the 'Six Little Dragons' reached a valuation of 20 billion yuan through multiple rounds of high - value financing last year. However, despite such high valuations, their profit data is rather dismal, and so far, no company has an annual revenue of 1 billion yuan.

From launching consumer - oriented chatbots last year to focusing on the B - side this year, the 'Six AI Little Dragons', with a single pre - training cost of three to four million US dollars, even Zhipu, which is the first to sprint for an IPO, according to Caixin Magazine, had a revenue of 300 million yuan and a loss of about 2 billion yuan in 2024.

According to The Information, in 2024, OpenAI's costs were mainly divided into three parts: inference costs, training costs, and labor costs, totaling about $8.5 billion. Comparing with OpenAI, the gap between the R & D expenses and commercialization revenue of the 'Six Little Dragons' is difficult to bridge in the short term.

The 'Six Little Dragons' that hope to leap over the dragon gate now urgently need to tell a new story of revenue growth to the outside world and demonstrate their self - financing ability.

Cost - cutting seems to be the first step in improving the data.

Yuezhi Anmian, which was quite aggressive in advertising and customer acquisition, with a monthly advertising expenditure of over 100 million yuan, has suspended large - scale advertising. Jieyue Xingchen recently reported the adjustment and reduction of its AI social tool 'Maopao Ya' business and is determined to cooperate with leading enterprises in the automotive and embodied - intelligence fields to develop vertical agents. Wei Wei, the former head of MiniMax's B - side business, was also reported to have left. Baichuan Intelligence laid off its B - side teams responsible for the finance, education, and other fields, aiming to concentrate resources on the core medical business.

Since it's difficult to tell a good story in the C - side market, the B - side has become a new breakthrough. However, as DeepSeek, Tongyi Qianwen, and ByteDance Doubao are competing to lower API prices, large enterprises that have accumulated loyal customers through cloud services in the mobile Internet era have launched "combined discounts on cloud services and large models" activities. There is even a possibility that the original B - side customers of the 'Six Little Dragons' may be poached by large enterprises.

"We can't compete with Alibaba and Baidu in API prices, and we don't have long - standing customers bound by cloud services. Last year, several companies almost subsidized the development of small - and medium - sized enterprise customers just to show business potential in the reports to investors," a marketing staff member of a large - model startup said.

In an interview with Intelligent Emergence, several employees of Baichuan Intelligence also said that since last year, many hospital departments have had indicators for AI implementation, and customers are willing to sign contracts. "One reason is for personal achievements, the second is to publish papers related to AI in the medical field, and the last is for the subsequent sales of medical products."

Taking Baichuan, which focuses on the B - side medical market, as an example, although it positioned itself in the medical field early and signed many B - side hospital customers during the window period, in mid - March, there were reports that Huawei was forming a healthcare business group. Currently, how to avoid direct competition with Huawei, the "B/G harvester", is a new topic that Baichuan is urgently discussing internally.

In addition, the 'Six Little Dragons' also have to face the question: what if the listing fails?

As a technology startup, Megvii, which once competed with SenseTime for the first - stock