Summary of Key Points
This news article focuses on the plummeting stock price of AI company MiniMax following its delisting, revealing a shift in the market's valuation logic for AI companies: once-regarded as newcomers that drove their market value through impressive narratives (such as MiniMax), these companies saw their stock prices halve due to issues such as weak model capabilities, lackluster consumer-facing businesses, and regulatory pressures. In contrast, established players like Tencent and Alibaba, which were once criticized for being “outdated” in AI, have actually implemented profitable AI services, though their value was previously overlooked. The ultimate conclusion is that newcomers lack tangible performance, while established companies struggle with insufficient growth momentum. Both groups face challenges, and market valuations are shifting from relying on hype to focusing on tangible results.
Detailed Analysis
1. How did MiniMax’s “promised greatness” evaporate?
MiniMax’s stock price dropped from 410 billion yuan to 90.8 billion yuan, not because of any mistakes, but due to the collapse of two key pillars supporting its market value:
- Overhyped model capabilities: The company claimed to be part of the “first-tier” all-modal AI teams, but third-party evaluations dispelled these claims—its models ranked ninth in Artificial Analysis and outside the top 50 in Chatbot Arena. More damningly, just a week after the launch of its new M3 model, it cut the price of its API services by half, prompting JPMorgan to downgrade its rating from “overweight” to “neutral” and reduce its target price by two-thirds.
- Unsustainable consumer-facing business: 60%-70% of MiniMax’s revenue came from emotional companionship products like Talkie. Although these products sounded promising at the time of listing, they revealed poor profitability, with a gross margin of only 4.7%, and monthly active users plummeted by 60% in the fourth quarter of last year. Additionally, new regulations on simulated human-like emotional interactions led to the immediate removal of related features from products like DouBao and QianWen, hitting MiniMax hard.
- Delisting exacerbated problems: At the time of delisting, only 3%-5% of its shares were publicly traded, meaning the 410-billion yuan market value was largely based on speculation. With 44.85% of its shares entering the market after delisting, the circulating supply increased tenfold, leading to a 20% drop in its stock price and a six-month low.
2. Are established companies really falling behind in the AI era?
Companies like Tencent and Alibaba, once criticized for being “old-fashioned,” actually have more substantial AI businesses:
- Alibaba: Its AI-related revenue from Alibaba Cloud exceeded 35.8 billion yuan in the 2026 fiscal year, showing triple-digit growth for 11 consecutive quarters—more than 70 times MiniMax’s annual revenue of 500 million yuan. Alibaba Cloud provides AI tools to businesses; while the companies using these tools may face risks, Alibaba itself generates stable profits from selling these services.
- Tencent: Its practical AI initiatives have paid off. Despite earlier criticisms of lagging in AI development, Tencent’s new Hy3 model led to a two-day increase in its stock price. This is because its model, with only 295 billion parameters (2-5 times smaller than others), performed on par with more expensive models and had lower inference costs. Moreover, by integrating it into the WorkBuddy office tool, users’ actual usage optimized the model, creating a positive cycle where better performance led to stronger models and greater product popularity.
3. Market valuation logic: shifting from hype to reality
In the past two years, the market believed that the more “exciting” the AI narrative, the higher the stock price. MiniMax’s claims of being a leader in all-modal AI and entering a new emotional companionship market helped it reach a market value of 410 billion yuan. However, its recent collapse has shown that even the best stories fail without profit and real capabilities. Conversely, the value of established companies lies in their tangible achievements—such as Alibaba’s growing AI revenue and Tencent’s successful integration of AI into everyday tools.
4. Both newcomers and established companies face challenges
- Newcomers’ challenges: Their promises haven’t translated into actual profits, and regulatory pressures are a major concern. Without performance improvements, their valuations will likely continue to decline.
- Established companies’ challenges: They need to convert their AI capabilities into sustained growth. Alibaba’s slow e-commerce growth and challenges in the food delivery market, Tencent’s mixed success with its Hy3 model, and Meituan’s lackluster AI efforts all highlight the difficulty of turning technological advantages into tangible business results.
In summary, the bubble in the AI industry is bursting, and the market is starting to judge companies based on their actual performance rather than empty promises. Both newcomers and established players must provide concrete evidence of their capabilities to gain investor confidence.