MiniMax can be seen as the primary beneficiary of OpenClaw's meteoric rise.

As AI companies and individual users rush to capitalize on the 'lobster farming' craze, MiniMax is already reaping substantial financial rewards.

According to Jiemian News, on March 10, MiniMax's stock price closed 22.37% higher, with its total market capitalization exceeding HK$380 billion. The previous day, its stock price had already surged by 23.77%.

This remarkable performance comes less than 60 days after its listing on the Hong Kong Stock Exchange, with the stock price having risen 6.4-fold and the market capitalization at one point surpassing that of established internet giants like Baidu and JD.com.

"The Pinduoduo of AI"

Many interpretations view this as a significant shake-up in the internet landscape, while others calmly point out that it is merely an AI illusion amplified by capital market sentiment.

Some even label MiniMax as the 'AI counterpart of Pinduoduo.'

In my opinion, comparing MiniMax to 'the Pinduoduo of AI' is not derogatory but rather an accurate depiction of its business model, growth trajectory, and operational logic, which differ from those of traditional AI behemoths.

The primary driver behind MiniMax's recent stock price surge is the popularity of OpenClaw (Lobster). Leveraging its low-barrier, high-value products, it has successfully capitalized on this trend.

As one of the official native large model suppliers endorsed by OpenClaw, MiniMax launched the cloud-based AI assistant MaxClaw, built on OpenClaw, on February 26. On March 9, MiniMax unveiled new 'Lobster' skills, introducing the Voice Maker voice model and Music Maker music model, further enhancing its multimodal capabilities.

This 'dimensionality reduction adaptation' strategy echoes Pinduoduo's early focus on lower-tier markets: instead of competing with traditional giants on heavy assets and full-scene coverage, it targets C-end users, transforming essential AI needs into accessible services for the masses, thereby expanding the market.

For a considerable time, traditional giants have pursued a 'big and comprehensive' approach, aiming to establish a full-stack technical route of 'AI cloud-model-chip.' The layout of leading AI firms also emphasizes full industry chain coverage, involving significant asset investments, relentless pursuit of core technologies, and targeting enterprise-level businesses that generate deterministic revenue to build technical barriers and stable commercial closed loops.

Entering the Agent era, the demand for 'AI that can get things done' has become paramount. AI giants have begun investing heavily in the user end, hoping to secure the AI2C super entrance.

AI newcomers like MiniMax face an 'asymmetric' or 'unequal' war of attrition. For instance, during the Lunar New Year AI battle, market investments exceeding RMB 10 billion from these giants posed a significant existential threat to them.

This has prompted them to adopt strategic avoidance, not engaging head-on with giants on general-purpose AI assistants but instead focusing limited resources such as funds and personnel on four major modalities: text, video, voice, and music.

This vertical model approach, with separate strikes but aiming for full-modality self-research, although initially questioned for dispersing R&D resources, has instead allowed MiniMax to focus on AI2C from the outset. It packages large model products into minimalist hosted versions, achieving lightweight, vertical breakthroughs on the C-end.

MiniMax's revenue structure attests to the effectiveness of its strategy. According to its first earnings report after listing, revenue from AI-native products for the C-end accounted for 67.2%, up 143.4% year-on-year; revenue from open platforms and enterprise services for the B-end accounted for 32.8%, up 197.8% year-on-year. This means that approximately two-thirds of MiniMax's revenue comes from C-end products.

This is a very astute approach. AI2C serves to attract users, bringing them in and validating products, technologies, and business models. Once validated, capabilities are then exported to high-margin B-end open platforms.

This also explains why MiniMax explicitly chose not to pursue general-purpose models from the outset.

Huxiu commented that MiniMax has built a 'full-modality layout, high-frequency iteration, and extreme cost-effectiveness' trifecta moat over four years, even catching up with overseas giants in multiple core sectors.

This strategy has two effects: First, it helps MiniMax bypass sensitive 'data walls' and reap the benefits of overseas markets. By 2025, its international revenue will account for 73.0%, serving a cumulative 236 million users across more than 200 countries and regions globally.

Second, MiniMax's cost-effectiveness has become its core competitiveness in the Agent era. OpenRouter data shows that in February 2026, the global weekly invocation volume of Chinese large models surpassed that of the United States for the first time, with MiniMax M2.5 ranking among the top five globally.

This coincides with the 'lobster farming' craze, and its representative Agent ecosystem is a veritable 'Token black hole,' restructuring the cost structure of AI Agents. In terms of input price, MiniMax M2.5 is only $0.3 per million Tokens, while Claude Opus 4.6 costs $5 during the same period; in terms of output price, MiniMax is $1.1, while Claude is as high as $25.

In other words, at the same capability level, MiniMax is 10 to 20 times cheaper. This extreme cost-effectiveness has helped it gain more market attention and users, almost solidifying its reputation as 'the Pinduoduo of AI.'

Breaking the AI Illusion

Objectively, it must be acknowledged that MiniMax possesses cutting-edge technology. For example, in the video generation field, its Conch series models have already ranked among the top tier globally; in the voice intelligence field, the Speech 2.6 model covers more than 200 countries and regions, becoming one of the most widely used AI voice models globally.

At the same time, as one of the few AI-native enterprises among listed companies, MiniMax's advantages and scarcity have been infinitely amplified in this 'lobster farming' craze, carrying excessive market expectations for future high growth.

Figure | Online Sources

Some analysis firms have stated that the company is at a critical juncture in the AI Agent industry's explosion, possessing both short-term high revenue growth and long-term ecological positioning value. As the number of installed users of OpenClaw grows, it may further drive significant growth in MiniMax's model Token volume and inference computing power and cloud demand.

However, from a long-term AI development perspective, this round of high market capitalization supported by traffic and sentiment resembles a bubble inflated by AI illusions rather than a reflection of true value. This does not mean that other AI firms are lagging or have taken the wrong path.

A gradually forming consensus is that 'lobster farming' is a temporary trend and may not be a long-term essential need for users.

I have also previously analyzed that this lobster is still 'raw,' with obvious shortcomings, far from the miraculous capabilities touted by some course sellers. Issues such as security vulnerabilities (system-level risks, data leaks), high usage costs (Token consumption, hardware thresholds), and technological instability (infinite loops, misoperations) are all problems that need to be addressed for OpenClaw's widespread adoption.

Especially at the security level, the Ministry of Industry and Information Technology promptly poured cold water, issuing an urgent reminder: 'Lobster,' under default or improper configurations, is highly susceptible to network attacks, information leaks, and other security issues. The ministry advised relevant units and users to thoroughly verify public network exposure, permission configurations, and credential management when deploying and applying it, closing unnecessary public network access and improving security mechanisms.

Once the hype around OpenClaw subsides and public novelty fades, this wave of casual traffic is likely to quickly dissipate. Without stable user stickiness and essential scenarios to support it, MiniMax's stock price and market capitalization face the risk of correction at any time.

A deeper issue is that MiniMax's business line is relatively single, and it has not yet established a commercialization closed loop. Before the profit model for AI Agents becomes clear, the pressure from its high market capitalization will be even greater.

Fundamentally, AI must first consider 'real demand.' MiniMax has accelerated the popularization of AI Agents and currently holds a market advantage, but it needs to focus more on deepening essential scenarios and refining its commercialization closed loop to convert technological advantages into stable revenue and cash flow.

MiniMax has already recognized this. After releasing its latest earnings report, its founder, Yan Junjie, anchored the company's strategy as 'moving from a large model company to an AI-era platform company.' He believes that platform companies in the internet era are traffic entrances, while platform companies in the AI era can define and drive new intelligent paradigms and enjoy paradigm dividends in products and commerce.

This is also an inevitable path for the entire AI industry: abandoning impetuous traffic hype, returning to technological original intention and user needs, refusing to blindly follow concepts, and focusing on technological implementation and value creation.

The market capitalization bubble brought about by this OpenClaw craze will eventually fade. Only tangible commercial value, technological barriers, and user essential needs can maximize AI's value and enable it to stand out in the AI race.

References:

Jiemian News: 'MiniMax's Stock Price Surges 6.4-Fold in Two Months After Listing'

Tang Chen's Classmate, 'OpenClaw Goes Viral, Challenging Tencent'