Anthropic has lobbed itself into the trillion-dollar IPO scrum as AI outfits keep torching cash for compute.
According to the Financial Times, the Claude maker said it had filed paperwork with the US Securities and Exchange Commission. That starts a process that could put it on the cocaine nose jobs of Wall Street within a few months, assuming investors still fancy cash-burning AI labs.
OpenAI, recently valued at $852 billion, is expected to file shortly. SpaceX published its IPO prospectus last month and is aiming for a listing at a $1.75 trillion valuation in the coming weeks.
Anthropic closed a $65 billion funding round last Thursday, valuing the outfit at $900 billion before the new cash was counted. The round brought in big public-market investors including Capital Group and Baillie Gifford, along with venture firms and strategic partners.
Multiple investors in the company reckon that will probably be Anthropic’s last private funding round. Anthropic has not yet set a price target for the listing, but its backers expect the IPO to value it at more than $1 trillion.
The company said the confidential filing “gives us the option to go public after the SEC completes its review. The proposed initial public offering will depend on market conditions and other factors.”
US tech companies often file IPO paperwork confidentially to keep touchy financial data hidden while the SEC has a look. That lets start-ups test investor demand, tweak documents and, if the market starts coughing blood, bin the whole idea without too much public embarrassment.
Anthropic’s valuation has jumped more than 15-fold in 15 months, after a March 2025 funding round valued it at about $60 billion. That surge has come from daft-looking revenue growth.
Anthropic’s annualised revenue, a start-up favourite based on recent performance rather than a full-year number, has risen fivefold since January. It passed $47 billion last month, helped by the company’s push into business customers.
The outfit has gained traction with coding model Claude Code and AI agent Claude Cowork. Anthropic’s AI model Mythos has rattled governments and financial institutions with its cybersecurity powers. Mythos is being gradually rolled out to US allies, including the EU.
Together, OpenAI, Anthropic and SpaceX-linked xAI have raised hundreds of billions of dollars while staying private. Most of that cash has gone into the computing power and infrastructure needed to train and run energy-guzzling AI models.
Anthropic has struck deals with Google, Broadcom and Amazon that could be worth hundreds of billions of dollars for future compute access.
The AI outfit also agreed last month to lease two huge data centre campuses from SpaceX in a deal worth $15 billion a year to the rocket maker.
As Anthropic, SpaceX and OpenAI head for public markets, their listings would rank among the biggest ever, alongside Saudi Aramco’s 2019 monster IPO.
