SpaceX, Anthropic and OpenAI are all planning an imminent IPO, promising a historic wave on Wall Street. Overview of the challenges they face as D-Day approaches.
Reduced to famine since 2022 and the rise in rates to curb post-Covid inflation, the IPO sector is on the verge of a spectacular rebound with the imminent arrival on the public markets of SpaceX, OpenAI (respectively the first and second most valuable unlisted companies in the world) and Anthropic. Each of them could reach a post-IPO valuation of a trillion dollars, while Elon Musk could become the first trillionaire in history.
The valuation of these three companies is closely correlated with the rise of AI, exclusively for OpenAI and Anthropic, and to some extent for SpaceX, which absorbed xAI in February. As shown by the recent (lost) lawsuit of Elon Musk against OpenAI, and the rivalry between Sam Altman and Dario Amodei, their respective histories and business models are closely linked.
SpaceX: the bet of New Space… or AI?
To all lords, all honor. SpaceX’s IPO is already historic and hyperbolic even before it takes place. Elon Musk is counting on a valuation of 2,000 billion dollars post-IPO, which would make it the largest IPO in history, ahead of that of Saudi Aramco, valued at 1,700 billion dollars following its entry on the Riyadh stock exchange in December 2019. SpaceX mainly intends to raise 80 billion dollars, much more than the 29 billion raised by Saudi Aramco in 2019. The IPO would of the boss of Tesla, the head of two of the most valuable American listed companies, an unprecedented situation in recent history. Let us add that it is hardly common to see a company specializing in the construction of rockets and satellites, in an American start-up market largely oriented towards software.
But SpaceX’s IPO is in reality also that of xAI, absorbed by SpaceX at the beginning of February. Elon Musk is therefore riding the rise of AI in addition to that of New Space and satellite internet to convince investors. SpaceX is also in the process of buying Cursor, one of the main specialists in vibe coding, which could give xAI a serious boost in AI, a boost that the company badly needs. Left behind on the power of its models by its rivals Google, OpenAI and Anthropic, the company suffered an operational loss of $6.4 billion, while most of its turnover comes from the social network X.
The IPO will therefore constitute a real-time test of Elon Musk’s course on AI, determining whether investors believe in xAI’s ability to carve out a place for itself against its competitors, and in the company’s ability to establish an ecosystem including model creation and application layer for businesses provided by Cursor. The latter constitutes a serious asset in this regard, as its growth is spectacular: it currently has 3 billion dollars in annualized turnover, compared to 2 in February, and is counting on 6 billion by the end of 2026.
The IPO will also test investors’ faith in the satellite internet market, for the moment still a niche service with ten million subscribers worldwide, but with strong commercial potential and constituting a powerful asset in terms of sovereignty.
Anthropic, the validation of a B2B strategy
Anthropic is in a strong position as it approaches its debut on the public markets. Since the release of Claude Code and Claude Work at the end of 2025, the company has established itself as the leader in agentic AI serving the democratization of AI in business. Driven by strong growth, Anthropic surpassed OpenAI in annualized revenue in early April.
And the company led by Dario Amodei has just achieved profitability for the first time, with an operating profit of $559 million in the second quarter, driven by a sharp increase in turnover ($10.9 billion, almost twice as much as in the first quarter). Anthropic’s revenue is now growing faster than Zoom’s during the pandemic. In addition to being likely to reassure investors, this news advantageously presents the company against its rival OpenAI, which for its part continues to suffer heavy losses And does not expect to be profitable before 2030.
The IPO thus constitutes in many respects a plebiscite on the professional agentic AI strategy on which Anthropic has bet, as opposed to the more general public one of OpenAI, a bet whose relevance the markets seem to validate for the moment.
This does not mean, however, that Anthropic has nothing to worry about in anticipation of its entry into the markets. First, the company has certainly achieved profitability by focusing on a more frugal model (fewer GPUs purchased, less investment in data centers) and by targeting a smaller audience of professionals rather than the mass of consumers. But it also achieved this result at the cost of restricting uses for users, of a financial arrangement with SpaceX which allowed it to temporarily obtain a discount on the use of the Colossus supercomputer. The company will sooner or later have to realize costly additional investments in the cloud to continue to fuel its growth. The company has also signed a series of new data center contracts in recent weeks in order to increase its capacities, notably with SpaceX.
In addition, if Dario Amodei’s relations with the White House have recently improved, investors will still have in mind the heated argument which pitted the company against the American government in March, and following which it was briefly ostracized by the American administration. The subject is thorny for Anthropic: on the one hand, the company is playing on its image as a white knight of AI against its rivals, including OpenAI, an image that too close proximity to the Trump administration and the American army could taint. On the other hand, an open war with it could deprive Anthropic of lucrative professional contracts and slow down one of its main growth levers.
OpenAI or the bet on growth at all costs
The issues raised by OpenAI’s IPO are almost the mirror image of those of Anthropic. Sam Altman’s company has in fact adopted a strategy radically different from that of its young rival. Where Anthropic strives to quickly achieve profitability by offering a product designed for adoption by businesses and by focusing on the sale of professional subscriptions, OpenAI has chosen to target the general public and quickly reach a critical mass of users, even if it means losing a lot of money, by planning to eventually become profitable thanks to advertising. A model that echoes the one that has underpinned the prosperity of giants like Google and Facebook.
It is also to catch up with Anthropic that OpenAI decided to accelerate its IPO. Its strong point is the popularity of its product among the general public: 920 million weekly users, which gives it a significant network effect and should convince investors that the company is not about to be ousted from the AI race. On the other hand, it only has 55 million paying subscribers, which, combined with its enormous investments to acquire more computing power, weighs on its financial results.
At the time of the IPO, investors will therefore judge its ability to convert other paying users, but also to generate more revenue through advertising. Despite difficult beginningsOpenAI is counting on 100 billion in turnover enabled by advertising by 2030. An ambitious figure, since it is half of what the formidable advertising machines of Google and Meta generate annually (respectively 214 and 196 billion dollars). Let’s see what investors think.
