SpaceX, OpenAI, and Anthropic represent the top forces in the space economy and general artificial intelligence respectively. Their listings are not only a feast for capital frenzy but also a battlefield for the game between value and speculation.
In the US stock market in 2026, it is standing at a historic crossroads. SpaceX, the absolute leader in commercial space, and OpenAI and Anthropic, the two oligarchs in generative AI, are intensively sprinting for IPOs, which constitute the three most eye – catching events in the global capital market this year. One of these three companies is targeting the trillion – dollar blue ocean of space infrastructure, while the other two are vying for the right to define general artificial intelligence. Their combined valuation may exceed $3 trillion, far surpassing any previous technology IPO combination.
When the fanatical capital pursuit meets Buffett’s cold metaphor of “a church with a casino attached”, and when the long – term value of hardcore technology collides fiercely with the bubble risk of short – term speculation, we need to penetrate the valuation fog, dissect the core competitiveness, business models, and industry impacts of the three companies. Moreover, in the game between value and speculation, we should explore the long – term trends of the commercial space and AI industries.
01 Core Portraits of the Three IPOs: Valuation, Selling Points, and Differentiated Patterns
SpaceX: The Trillion – Dollar Overlord of Space Infrastructure, Valued at $1.75 Trillion
As the core of Musk’s business empire, SpaceX is scheduled to list on the NASDAQ in June 2026, with the code “Project Apex”. It aims for a valuation of $1.75 trillion and plans to raise $50 – 75 billion, which will surpass Saudi Aramco to become the largest – scale IPO in human history. Its valuation logic has long deviated from the traditional manufacturing framework. The core selling points lie in the three – dimensional monopoly barrier of “hardware + service + data” and the absolute voice in the space economy.
From a fundamental perspective, SpaceX’s cash – cow business, Starlink, has built an irreplicable moat: it has deployed over 9,500 low – orbit satellites in total, covering 99% of the global population, with over 10 million users. In 2025, its revenue was approximately $12.3 billion, accounting for 70% – 80% of the company’s total revenue. The revenue structure presents a golden combination of “consumer – level subscriptions + enterprise – level services + government defense orders”: individual users pay $120 per month, while enterprise/maritime/aviation customers pay over $5,000 per month. The Starshield project in cooperation with the US Department of Defense provides long – term and stable large – scale contracts. Quilty Space predicts that Starlink’s revenue will reach $20 billion in 2026, with an EBITDA of approximately $14 billion, and the cash flow will remain positive.
In addition to Starlink, SpaceX’s rocket launch business (Falcon 9 and Starship) occupies over 70% of the global commercial launch market. As the next – generation reusable heavy – lift rocket, the Starship can carry 60 satellites at a time, which will significantly reduce the cost of space deployment and provide hardware support for future visions such as space tourism, lunar bases, and Mars colonization. What is more imaginable is the potential value of space data assets. The low – orbit satellite network is not only a communication infrastructure but also a central hub for the collection and processing of space data. In the future, it may give rise to new business models such as data trading and space cloud computing.
SpaceX’s differentiated advantage lies in its irreplaceable technological barrier, cost – control ability, and ecological closed – loop: the reusable rocket technology has reduced the launch cost by 90%. The spectrum and orbital resources of Starlink are naturally scarce, and Musk’s personal vision and execution ability guarantee long – term technological investment. Its valuation premium is essentially the pricing of its status as the “global monopolist of space infrastructure”, rather than a simple valuation based on financial statements.
OpenAI: The Traffic King of Generative AI, Valued at $852 Billion
OpenAI is scheduled to list in the fourth quarter of 2026, with a valuation of $852 billion. It plans to raise over $60 billion and is the unicorn with the highest valuation in the AI track. From a non – profit organization in 2015 to a current commercial giant, OpenAI’s core selling points are its globally leading large – model technology, hundreds of millions of user scale, and the ultimate narrative of “AGI (General Artificial Intelligence)”.
At the product level, OpenAI has built a full – stack AI product matrix centered around GPT – 4o: ChatGPT on the C – side has over 900 million weekly active users, making it the AI application with the largest user scale globally; on the B – side, it covers enterprise API interfaces, Codex code – generation tools, AtlasAI search, etc., covering all scenarios such as finance, healthcare, law, and programming. In 2025, its revenue was approximately $24 billion, of which 73% came from ChatGPT personal subscriptions, and enterprise API revenue accounted for approximately 40%. In terms of equity structure, Microsoft holds 27% of the economic interests, making it the largest external shareholder and computing – power provider. SoftBank, Thrive Capital, etc. are the core investors, with a cumulative financing of over $64.1 billion.
In terms of business model, OpenAI adopts a diversified monetization strategy of “freemium + subscription + API call”: it acquires a large number of C – side users through the free version of ChatGPT, realizes C – side monetization through the Plus subscription service at $20 per month, and at the same time opens API interfaces to enterprises and developers, charging according to the call volume. Its core advantages lie in the ecological barrier brought by the first – mover advantage, the leadership in multimodal technology, and brand recognition: the GPT series of models were the first to achieve commercial implementation, accumulating a large amount of user data and a developer ecosystem, forming a positive cycle of “more data – stronger model – more users”.
However, OpenAI’s shortcomings are also significant: the profit dilemma of increasing revenue but not profit, the low payment rate of C – side users, and the high computing – power cost. In 2025, the net loss was approximately $14 billion, and the total debt reached $436 billion. Most of the 900 million weekly active users are free or low – paying users, and the inference cost remains high. Its valuation highly depends on the AGI narrative and user – growth expectations, and its commercial monetization ability has not been fully verified.
Anthropic: The Pragmatist in Enterprise – Level AI, Valued at $380 Billion
Anthropic was founded in 2021 by the former core team of OpenAI. It is scheduled to list in October 2026, with a valuation of over $380 billion. It is the most direct and powerful competitor to OpenAI. Different from OpenAI’s C – side traffic route, Anthropic’s core selling points are the precise positioning in the enterprise – level market, the safe and controllable technology route, and the high – margin business model.
At the product level, Anthropic focuses on the Claude series of models, emphasizing “safety, reliability, and interpretability” and targeting enterprise – level scenarios: Claude 3 (Opus/Sonnet/Haiku) specializes in long – text processing (supporting 200k context), enterprise – level security compliance, and high stability. Claude Code has become the preferred coding tool for developers. In 2025, the annualized revenue exceeded $30 billion, and 80% of the revenue came from enterprise – side customers. The paying customers are mainly medium – and large – sized enterprises in industries such as finance, technology, and healthcare, with high customer unit prices and strong stickiness.
In terms of business model, Anthropic adopts a pure enterprise API – driven model. 85% of its revenue comes from API calls by developers and enterprises, charging according to the number of tokens. The marginal cost is extremely low, and the scale effect is significant. Its differentiated advantage lies in the depth of enterprise – level services, the security and compliance barrier, and the efficient monetization ability: the team comes from the core of OpenAI, and its technical strength is comparable to GPT – 4; it focuses on the enterprise market, avoiding the red ocean of C – side traffic, with a customer renewal rate of over 90%; it emphasizes AI security, meeting the global enterprise data compliance requirements.
The core of the competition with OpenAI is the route competition between the “platform empire” and the “expert blade”: OpenAI pursues breadth, trying to build an all – inclusive AI super – application; Anthropic focuses on depth, deeply cultivating enterprise – level scenarios. Among investors, OpenAI is backed by capital giants such as Microsoft and SoftBank, and its valuation is more inclined to growth and narrative; Anthropic has received investments from enterprise giants such as Google and Amazon, and its valuation pays more attention to cash flow and profit expectations.
02 The Duel of the Two AI Giants: Comprehensive Competition in Products, Business Models, and Investors
The competition between OpenAI and Anthropic is the ultimate battle in the generative AI track, covering all – dimensional confrontations in products, business models, technology routes, and investor ecosystems. In essence, it is a collision of two AI commercialization philosophies.
At the product level, OpenAI focuses on being “all – around”, using ChatGPT as an entry point to integrate functions such as dialogue, search, programming, and multimodal generation, pursuing user scale and the breadth of scenario coverage; Anthropic focuses on being “professional”, centering around Claude, and targeting enterprise long – text processing, security compliance, high stability, and other scenarios, pursuing service depth and customer value. In terms of user structure, OpenAI mainly has C – side individual users, and the proportion of paying users among the 900 million weekly active users is less than 10%; Anthropic mainly has B – side enterprise customers, and over 90% of its revenue comes from 300,000 enterprise customers, with a customer unit price more than three times that of OpenAI’s enterprise customers.
At the business – model level, OpenAI is a “traffic – monetization type”, relying on C – side subscriptions and advertising, with large revenue fluctuations and high marginal costs. In 2025, the net loss rate exceeded 50%; Anthropic is an “enterprise – service type”, relying on API calls and customized services, with stable revenue and low marginal costs. In 2025, the gross profit margin reached 75%, and it has achieved single – quarter profitability. The core difference is: OpenAI exchanges a large number of users for valuation, while Anthropic exchanges stable cash flow for valuation. The former bets on long – term AGI, while the latter focuses on short – term profit realization.
Among investors, OpenAI’s shareholders are mainly growth – oriented capital. Microsoft, SoftBank, Thrive Capital, etc. pay more attention to user growth and technological barriers and tolerate long – term losses. The valuation premium mainly comes from narrative expectations; Anthropic’s shareholders are mainly industrial capital. Google, Amazon, Salesforce, etc. pay more attention to enterprise – ecosystem synergy and cash – flow returns and require a clear profit path. Its valuation is closer to the fundamentals. This difference in investor structure determines the market performance of the two companies after listing: OpenAI’s stock price may fluctuate more, relying on sentiment and narrative; Anthropic’s stock price is more stable, relying on performance realization.
The key to winning this competition lies in the balance between the speed of commercial implementation and technological iteration: OpenAI needs to solve the problems of low C – side payment rate and high cost and accelerate penetration into the enterprise side; Anthropic needs to expand its technological influence and moderately expand the C – side or developer ecosystem while maintaining its enterprise advantages. Regardless of the outcome, the duel between the two AI giants will accelerate the maturity of the industry and push AI from “telling stories” to “making real money”.
03 SpaceX Leads a New Cycle in the US Stock Market: Is It a Bubble Frenzy or a Value Re – evaluation?
SpaceX’s trillion – dollar valuation IPO is not only a listing event for a single company but is also regarded as the “new ceiling” for US technology stocks, marking the reconstruction of the pricing logic of the capital market for hardcore technology and long – term – oriented assets. However, under Buffett’s metaphor of “a casino and a church”, behind this frenzy lies an intertwining of value re – evaluation and bubble risk.
From a positive perspective, SpaceX’s listing will kick off the “hardcore era” of US technology IPOs: in the past, most US technology giants (Apple, Microsoft, Google) relied on software and Internet services, with light assets and quick monetization; while commercial space represented by SpaceX is hardcore technology with heavy assets, long cycles, and high barriers, requiring continuous huge investments, and the return cycle is as long as 10 – 20 years. Its trillion – dollar valuation means that the capital market is willing to pay for the long – term vision of “human – civilization – level” and recognizes the long – term value of frontier technologies such as the space economy and AI, pushing the US stock market from the “Internet bubble” to the stage of “hardcore technology value re – evaluation”.
Meanwhile, SpaceX’s listing will drive the value re – evaluation of the entire commercial space industry chain: companies in sub – tracks such as rocket launch, satellite manufacturing, space communication, and aerospace materials will receive capital attention and valuation premiums, accelerating technological iteration and commercial implementation. In addition, after SpaceX is included in the S&P 500 index, it will attract tens of billions of dollars in passive funds, further consolidating its market position and forming a positive cycle of “giant listing – capital inflow – valuation increase – refinancing expansion”.
From a risk perspective, the valuation bubble of SpaceX cannot be ignored: a valuation of $1.75 trillion corresponds to a revenue of $12.3 billion in 2025, with a price – to – sales ratio as high as 142 times, far higher than that of mature technology giants such as Tesla (price – to – sales ratio of 15 times) and Apple (price – to – sales ratio of 7 times). Even if the revenue reaches $20 billion in 2026, the price – to – sales ratio will still exceed 80 times, over – estimating the growth expectations for more than 10 years in the future. This valuation highly depends on Musk’s personal ability, the user growth of Starlink, and the realization of the space – economy vision. Once there are technological bottlenecks, growth falls short of expectations, or management turmoil occurs, the valuation will face a significant correction risk.
A deeper concern is that the high valuations of SpaceX and AI giants reflect the “excessive gambling nature” of the current US stock market – as Buffett said, the market has become a “church with a casino attached”. The “church” of value investment is being wrapped up by the “casino” of short – term speculation. Investors chase narratives and ignore fundamentals, resulting in asset pricing deviating significantly from actual value. The valuations of companies such as SpaceX and OpenAI are more of “expected pricing” rather than “performance pricing”. Once the market sentiment reverses, the bursting of the bubble will trigger a chain reaction.
04 The Future of Technology under the “Church and Casino”: The Dialectics of Pessimism and Optimism
At the 2026 Berkshire Hathaway Annual Shareholders Meeting, Buffett’s metaphor of “a church with a casino attached” accurately summarized the core contradiction in the current global capital market: on one side is the adherence to long – term value (the church), and on the other side is the frenzy of short – term speculation (the casino). For the two tracks of commercial space (SpaceX) and AI (OpenAI, Anthropic), pessimism and optimism are not absolutely opposite but depend on the time dimension and value – judgment criteria from which we view the industries.
From a short – term (1 – 3 years) pessimistic perspective, both tracks face triple pressures of valuation bubbles, profit dilemmas, and technological bottlenecks. In the AI track, OpenAI incurs an annual loss of $14 billion and has high debts. Although Anthropic is profitable, its scale is limited. Both companies need to continuously invest huge computing – power costs, and the difficulty of commercial monetization far exceeds expectations. The large – model technology is approaching the “ceiling”, the performance improvement brought by parameter growth is diminishing marginally, while the computing – power cost is increasing exponentially, and the uncertainty of technological breakthroughs has increased significantly. In the commercial space track, SpaceX’s valuation over – estimates future growth. The user growth of Starlink has slowed down, and competition has intensified. The rocket – launch business faces challenges from opponents such as Blue Origin and Virgin Galactic. The commercial implementation of the space economy (space tourism, Mars colonization) is still far away. More importantly, the current market has an excessive gambling nature, and the valuations of AI and aerospace concept stocks are generally inflated. Once the Federal Reserve raises