Author: Think AI, Aaron

Less than half a year after its listing, Minimax's wealth creation myth has begun to collapse. (Related reading: MiniMax: A youth from a county in Henan and his 300 billion)

This January, MiniMax partnered with Zhizhu to list on the Hong Kong Stock Exchange, becoming one of the world's first publicly listed native companies for large models, once referred to as the AI twin stars of Hong Kong stocks.

However, currently, MiniMax's stock price has plummeted by 70% from its peak, with its market capitalization dropping from over 400 billion to more than 130 billion.

This means that in just 5 months, MiniMax has evaporated over 300 billion in market value, equating to the loss of a Baidu.

However, the most affected may not be ordinary investors, but the company's internal employees.

The most talked-about aspect of MiniMax is still the story of everyone becoming rich. The company has a total of 385 full-time employees, but 392 people hold shares, including interns and former employees, truly representing full participation in stock ownership, and it is currently the most insane share incentive plan among domestic companies.

The average value of shares held by employees was highest in March when MiniMax surged, reaching over 42 million, but with the plummeting stock price, it is now valued at over 12 million, equivalent to an average unrealized loss of 30 million per employee.

How do MiniMax employees view the drastic depreciation of their shares? As one of the earliest commercialized companies among China's six little AI tigers, with the highest overseas revenue and the most willing capital market, how will MiniMax navigate out of this dilemma after the stock price crash?

The Wealth Creation and Retracement of All-Staff Stock Ownership

It must be said that worrying about MiniMax employees facing an average unrealized loss of 30 million might be excessive, after all, the shares they hold are still worth over 12 million.

Many people, upon seeing this number, may wonder if the editor miscalculated. No need for doubt; here’s a brief introduction.

MiniMax's employee stock ownership plan has always been a benchmark in the domestic AI industry.

Unlike most tech companies that only grant stock options to core personnel, MiniMax has almost achieved full employee coverage.

According to 36Kr, including former employees, there are as many as 392 people in MiniMax's employee stock ownership plan. Among them, 363 employees hold shares, accounting for about 3.41% of the total share capital issued after the IPO, currently valued at nearly 4.7 billion HKD.

The core of full employee allocation lies in two points: the company itself has a small number of employees, and AI talent is currently in great demand—insufficient incentives lead to employees being quickly poached by competitors.

In terms of allocation rules, stock options are tilted towards technology and core personnel, but frontline employees can also receive substantial shares. The lowest exercise price is merely 0.002 USD/share, almost free distribution, with the binding strength unmatched among domestic large model companies.

The highest stock price of MiniMax was 1330 HKD, now priced at 435, a decline of 67.3%, close to seven tenths; the book value has indeed suffered significant losses.

However, all employee shares are still in a lock-up period, which will gradually be lifted starting January 2027. It is yet to be determined how much of this wealth roller coaster will ultimately be realized.

Minimax's Dilemma

Having discussed employee benefits, let's return to the company itself.

Many people may know about Doubao and Deepseek, but are less familiar with MiniMax. This is not surprising, as 70% of minimax's market comes from overseas, with a relatively small user base domestically.

MiniMax has already launched some quite remarkable products overseas, with a high degree of commercialization, with C-end applications like alkie reaching 210 million users globally, particularly favored by young customers overseas.

The video generation model from Hai Luo also ranks among the top tier. MiniMax Audio's speech synthesis and MiniMax Agent, which provides services to enterprises and developers, have received positive feedback.

Moreover, overseas developer call volumes often dominate the Openrouter rankings. Yet, this company has still experienced a crash; why is that?

Firstly, there is the impending wave of lock-up stock releases.

After listing on the Hong Kong Stock Exchange in January this year, most of MiniMax's shares will have a lock-up period of 6 months, with a large-scale release expected in July, accounting for over 65% of the total share capital.

Early investors like Sequoia, Hillhouse, and Mihayou have already seen returns of 4-5 times, and the expectation of liquidating shares after the lock-up has been a core incentive for funds to exit early.

Additionally, as many model APIs announce price cuts, MiniMax has raised interface prices, leading some small and medium developers to turn to competitors like DeepSeek, raising direct concerns about its commercialization stability.

A more insidious dilemma is that despite MiniMax having such a broad domestic market, it has not secured significant shares in both B-end and C-end markets. The lack of a local market base to hedge overseas risks has left its valuation story consistently lacking support from domestic user consensus.

As the capital market and AI application directions begin to shift towards Vibe Coding, Agents, and enterprise automation, large model companies with stronger B-end capabilities are favored by investors, while MiniMax has long leaned towards the C-end, raising doubts about its profitability and potential for growth.

Furthermore, Deepseek has also completed substantial financing, while OpenAi and Anthropic are about to go public, increasing the number of AI investment targets globally, leading to funds flowing away from MiniMax.

What Should Minimax Do?

When compared to other domestic large model players, MiniMax's position is actually quite delicate.

Compared to Doubao, what MiniMax lacks is a super entry point.

Doubao is backed by ByteDance, with national-level traffic platforms like Douyin, Jianying, and Toutiao, which can directly deliver AI features to hundreds of millions of users. Although MiniMax has Xinye and Hai Luo AI, it is more about vertical products, making it difficult to establish a nationwide recognition comparable to Doubao.

In comparison to DeepSeek, MiniMax lacks the technical breakthrough label.

DeepSeek has successfully positioned itself as a focal point in the global AI industry through low costs, open-source, and inference capabilities. MiniMax is also emphasizing open-source, long context, multi-modal, and Agents, but currently, the consensus in the B-end is insufficient.

MiniMax's advantages lie in overseas users, multi-modal products, and early commercialization capabilities. However, the capital urgently needs to see MiniMax's higher monetization ability.

To navigate out of the dilemma, MiniMax needs to focus on three areas.

Firstly, it is essential to rebuild developer trust. While the API price increase has been criticized, it is more important that users do not feel that the rules can change at any moment. In the future, large model companies will compete not only on parameters and rankings but also on ecology, reputation, and developer stickiness.

Secondly, it is necessary to truly turn M3 into a B-end product.

Lastly, MiniMax must increase the proportion of B-end and platform-type revenue, making the revenue structure more stable.

The 70% drop in MiniMax's stock price does not mean the AI story has ended, but rather the market has started to move away from FOMO, focusing on imagination, and is transitioning towards more realistic implementations.

However, for early employees, regardless of the outcome, they have already achieved financial freedom.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。