Lobster Wealth Creation! MiniMax's stock price hits a record high, with a market value approaching $400 billion.

Questioning how open-source project OpenClaw is rewriting the rules of competition in the AI industry?

A wave sparked by an overseas open-source project has propelled a newly established AI company into a two-month stock surge of nearly 200% in the Hong Kong stock market.

While the AI industry is still debating the limits of parameters and computing power, the capital market has already provided a new valuation perspective with real money.

MiniMax is arguably the biggest beneficiary of OpenClaw’s breakout.

As AI vendors and individual users scramble to get a piece of the “lobster” craze, MiniMax has already achieved significant gains on its books.

Looking back at this wealth-creation movement triggered by lobsters, it has not only rewritten the ranking of corporate market values but also redefined the competitive logic of the AI industry. The industry is shifting from a focus on technological worship to a pragmatic path of commercial implementation.

01 Lobster Beneficiaries: An Unexpected Capital Frenzy

The popularity of the OpenClaw open-source project has put MiniMax at the center of the industry spotlight. This AI agent framework, nicknamed “Lobster” by developers, with its ability to autonomously perform tasks and efficiently interface with underlying models, quickly swept through the global developer community. MiniMax is the core beneficiary of this wave.

In just over two months since listing on the Hong Kong Stock Exchange, MiniMax’s stock price has increased fivefold, with its market value surpassing that of established internet giants like Baidu and JD.com at times. This new AI star has achieved in two months what traditionally took decades for other companies.

The market frenzy is backed by solid business growth. According to data from the OpenRouter platform, in February 2026, MiniMax’s M2.5 model ranked first in call volume for three consecutive weeks, with a peak weekly call count of 3.07 trillion tokens, surpassing Kimi, Zhispur, and DeepSeek combined.

Public information shows that MiniMax has transformed from a simple model provider into a core infrastructure for the AI era, with developers in over 200 countries using its models, and overseas revenue accounting for over 70%.

The recent stock price rally is driven by the popularity of OpenClaw (Lobster). Its low-threshold, high-cost-performance products have successfully captured this wave of traffic.

This wealth-creation movement has led to a reassessment of the value of the entire AI track. MiniMax and Zhispur’s market caps are rising in tandem, with recent private funding rounds valuing the company at $18 billion.

A phrase has begun circulating in the industry: On the wave of opportunity, choosing the right track is more important than just R&D.

Looking deeper, in the business world, there are no unprovoked viral hits; all outbreaks are the result of long-term strategic accumulation. MiniMax’s rise may seem accidental, driven by the lobster concept, but in reality, it is the inevitable result of precisely positioning in the intelligent agent track and坚持全模态技术布局.

While the industry is still chasing the illusion of peak intelligence, MiniMax has quietly rooted itself in real-world applications, ultimately transforming from a startup into an industry giant when the opportunity arose. This confirms the principle that the ultimate value of technology lies in solving real problems, not just in parameter leaderboards in labs.

02 Giants Surround, Challenges Remain in the Spotlight

Standing at a market cap of 380 billion yuan, every move by MiniMax is under the spotlight, and the accompanying reality is full-scale containment by domestic and international giants.

The AI industry has always been a zero-sum game. When a startup breaks the existing pattern, it inevitably provokes retaliation from leading companies. This seemingly glorious breakout is actually fraught with danger.

In the domestic market, competition among the top tier has already intensified. Zhispur AI, leveraging Tsinghua University’s background, has deep roots in government and scientific research markets, securing advantages in the sovereign AI track with its safe and controllable features; Yuezhianan relies on long-text processing technology, gaining high popularity among C-end users. After releasing the K2.5 model, its revenue in less than 20 days exceeded the total for all of 2025. Its Kimi Claw can be deployed with one click, allowing users to access OpenClaw functions without hardware, servers, or coding.

DeepSeek, with its open-source strategy and cost-effective offerings, has established a foothold in the global developer community, maintaining top-tier performance among open-source models worldwide. These competitors are either well-funded, possess unique technology, or focus deeply on specific scenarios—each formidable.

On the overseas front, MiniMax faces pressure from giants like OpenAI and Google. These companies have stronger capital, larger datasets, and more mature ecosystems, dominating the high-end model market.

Financially, the company suffered a loss of $1.872 billion last year, mainly due to a $1.59 billion fair value loss on preferred shares. This is a “technical cost” of the company’s soaring valuation reflected in financial statements. Currently, preferred shares have been automatically converted into common stock, making this non-cash loss a thing of the past.

However, under the pressure of industry giants, challenges remain. Some analysts believe the company is at a critical point in the AI Agent industry explosion, with both short-term rapid revenue growth and long-term ecosystem positioning potential. As OpenClaw’s user base grows, it could further boost MiniMax’s token volume, inference computing power, and cloud deployment needs.

From a long-term AI development perspective, “raising lobsters” is a passing trend; it may not be a long-term user demand.

In other words, the moment of glory often harbors crises. Industry competition is never a sprint but a marathon without a finish line.

Historically, short-term market cap leadership does not guarantee safety. In an industry where iteration speeds are measured in months, any complacency can lead to being overtaken.

MiniMax’s challenge is no longer just about catching the wave but about holding ground amid giants and turning short-term hype into long-term core competitiveness.

On a broader scale, this reveals that the rise of startups is a journey of both opportunity and adversity. Behind the glamour are countless unknown challenges.

03 From Peak Intelligence Competition to Cost-Performance Race

From an industry perspective, the lobster craze signals a fundamental shift in AI competition logic.

In the past, the industry obsessed over trillion-parameter models and top evaluation scores, competing for peak intelligence. Now, most companies are focusing on one key word—cost-performance. This shift has transformed AI from a capital-driven concept into a practical, inclusive industry.

Cost-performance has become a core competitive advantage driven by real-world demand. According to the latest industry data in 2026, global mainstream large models show a four-tier price differentiation for tokens. Domestic models in China outperform international models in Chinese scenarios, with costs only 1/5 to 1/20 of top overseas models.

OpenRouter platform data shows that in February 2026, Chinese models accounted for 61% of global token consumption, surpassing the US for the first time. MiniMax M2.5, Zhispur GLM-5, and Kimi K2.5 ranked top in call volume. Behind this data is the scale effect brought by cost-performance: as AI usage costs plummet, countless vertical scenarios are activated—from enterprise digital employees to personal creative assistants, from code generation to content creation—AI is truly integrated into daily life.

Technological advances are also empowering cost-performance. Breakthroughs in MoE architecture, FP8 quantization, and all-optical interconnects have significantly improved inference efficiency, reducing computing costs. Low-cost green electricity contracts in western China, at just 0.13-0.3 yuan per kWh, turn energy cost advantages into model pricing advantages, enabling domestically developed models to compete globally on a lower-cost basis.

This has led to industry consensus: peak intelligence is just the foundation; solving more real-world problems at lower costs is the true core barrier for AI companies.

The shift from parameter competition to cost-performance race marks industry maturity. Once distant artificial intelligence has become a basic utility like electricity, representing a shift in business logic and a return to technological value. MiniMax’s rise exemplifies this trend—using extreme cost-performance to open global markets and guiding Chinese AI companies on their development path.

Looking back at the lobster wealth myth, it seems accidental but is actually an inevitable outcome of AI industry evolution. MiniMax’s soaring valuation is not just capital hype but market recognition of the value of real-world deployment and cost-performance competition.

Yet, beneath the glory lie challenges. The celebration is tempered by the need for calm. This industry transformation teaches us that AI’s future is never about a few giants’ technological frenzy but about grounded application and serving the masses.

From suppression to frenzy, from peak intelligence to cost-performance competition, the upheaval in AI is merely a reaffirmation of business principles. The companies that truly go far are those that understand the market, uphold value, and keep creating. The lobster craze will fade, but the trend of AI democratization and practicality is irreversible. The next wealth myth will be born on tracks that truly solve problems.