Anthropic backers eye $2 trillion IPO, topping SpaceX

Anthropic's IPO Path and the $2 Trillion Valuation Talk

Six major investors in Anthropic expect the artificial intelligence lab to go public in the coming months at a valuation of roughly $2 trillion (€1.73 trillion), the Financial Times reported Thursday, citing anonymous sources. That level would be twice the figure discussed informally inside the company in recent months and about $200 billion above the valuation achieved by SpaceX when it debuted on June 12.

The enthusiasm is tied to Claude. Anthropic co-founder Dario Amodei said at a May conference that the company's annualized revenue — the projection of a full year's sales from a shorter period — grew 80-fold in the first quarter of 2026. Investors surveyed by the FT expect another tenfold increase across the second half of the year, which would leave Anthropic with annualized revenue of about $110 billion (€95.3 billion) by the end of 2026.

That performance has been driven by new Anthropic models that can reproduce complex programming functions without code and handle office tasks such as financial planning or legal assistance. The company, led by siblings Dario and Daniela Amodei, reportedly estimates it closed the second quarter with its first profit, although quarterly results have not yet been published. One of the more optimistic investors told the FT that an 800% annual growth rate could justify a $3 trillion valuation, about 30 times revenue.

The same day, Bloomberg reported — also citing anonymous sources — that Anthropic is in talks to acquire AI start-up Decart AI for around $6 billion (€5.2 billion). Decart builds 'world models' that aim to simulate the physical world and software that reduces AI training costs by helping chips work more efficiently; a person familiar with the discussions said that technology could let Anthropic's existing infrastructure absorb more demand.

What a $2 Trillion Anthropic Would Mean for AI Monetization

The Claude Sales Bet Behind the Valuation

The bull case cited by Anthropic's investors is a claim that the AI platform is converting into stable, large-scale enterprise revenue. The hardest supporting data point is Dario Amodei's public statement that annualized revenue rose 80-fold in the first quarter of 2026; however, annualized figures extrapolate a short period across a whole year, making them particularly sensitive to a single quarter of acceleration. The FT's surveyed investors then go a step further, projecting a tenfold increase in the second half and an exit rate near $110 billion. At a $2 trillion IPO valuation, the market would be paying roughly 18 times that projected annualized revenue. That is steep by conventional software standards, but the more aggressive $3 trillion scenario — about 27 times revenue — rests on an 800% growth rate continuing without large margin erosion or customer churn. None of these figures has been confirmed by published financial statements.

The Decart Deal Is a Cost-Side Play, Not Just Another Model Purchase

Bloomberg's report of talks to buy Decart AI for about $6 billion may look like a talent or research acquisition, but the strategic logic is about capacity economics. Decart's software can reduce the cost of AI training by making chips operate more efficiently, which directly addresses one of the largest constraints on AI profitability: the capital and compute required to meet demand. Its 'world models' are a longer-horizon research effort, but the near-term value would be to let Anthropic serve more customers on its existing infrastructure at lower unit cost. That matters because the most credible threat to a $2 trillion valuation is not weak demand for Claude; it is the cost of running a rapidly expanding model business and the amount of new compute that would otherwise be required. The deal is not confirmed, and no purchase agreement had been announced at the time of reporting.

What a Profitable Quarter Would Prove — and What It Wouldn't

Anthropic reportedly estimates that it closed the second quarter of 2026 with its first profit. If confirmed, that would be a symbolic and financial turning point: it would give the company a concrete answer to the market's broad doubt that AI labs can monetize their products. But a single profitable quarter does not settle the question of structural profitability. The valuation debate depends on whether high growth and improving unit economics can coexist while Anthropic continues to spend on models, chips and acquisitions. The fact that results have not been published means the profit claim should still be read as an expectation rather than a disclosure.

Reading the SpaceX Comparison Carefully

SpaceX is a benchmark here, not a competitor. Its June 12 debut gave public markets a reference point for what a large, capital-intensive technology company can command. Anthropic's expected valuation of roughly $200 billion above SpaceX's debut would signal that investors are willing to apply a different multiplier to AI platform economics than to aerospace and launch services. That does not mean Anthropic's business has been proved stronger; it means the pricing logic is different. The more meaningful comparison for Anthropic is with the broader field of AI labs competing for enterprise contracts. The optimism among existing investors is a strong signal of their confidence, but external investors will need published financials to verify the trajectory before any IPO can deliver the reported valuation.

The Pre-IPO Signals for Anthropic's Investors, Customers and Rivals

For the people and companies directly exposed to Anthropic's trajectory, the actionable points follow from the few hard data points available:

• Anthropic's management and underwriters: Publish the unaudited Q2 2026 results before the IPO roadshow. The claim of a first profit is the strongest weapon against AI monetization skepticism, but it carries no evidentiary weight until the formal release.
• Prospective IPO investors: Ask for the quarter-by-quarter detail behind the 80-fold Q1 annualized revenue growth and the FT's surveyed expectation of roughly $110 billion in annualized revenue by end-2026. A $2 trillion valuation implies about 18 times that projected revenue; the $3 trillion scenario implies 800% growth persists, which is the central assumption to stress-test.
• Enterprise Claude customers: Treat the reported Decart AI talks as a signal of future cost and capacity improvements, not a committed product change. Decart's training-cost reduction software could improve Anthropic's pricing power and uptime, but the approximately $6 billion deal remains unconfirmed.
• Competing AI labs and their investors: Use Anthropic's reported profitability estimate and $110 billion run-rate target as the new benchmark for enterprise AI monetization. If those figures survive disclosure, they will raise the bar for how the market judges AI revenue quality across the sector.