[AINews] Anthropic files for $2T IPO in 2 weeks: 1400% growth rate, 48% of revenue through cloud partners, 85% margin, $1B op profit in Q2, $134B cash in bank and $10B lifetime burn
Latent Space swyx .ai ● Covered by 25 sources
Anthropic filed to go public and says it’s already profitable on an operating basis. The weird part: huge growth, huge cash, and huge compute bills all at once.
Based on reporting by Latent Space, swyx .ai — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Anthropic filed its S-1 on Monday morning and plans to list on Nasdaq under the ticker ANT. The document is the kind of filing that makes even seasoned AI watchers stop and squint: a lot of growth, a lot of cash, and a lot of obligations that run deep into the next decade.
Revenue rose from $41 million in 2023 to $16.3 billion in the first half of 2026. Second-quarter revenue was $11.6 billion, up 1,368% from a year earlier. More importantly, the company says it reached operating profitability in the quarter, with operating profit of $974 million, while first-half operating cash flow came in at plus $3.9 billion.
That profitability does not come from magic. A huge share of the business still runs through cloud partners: 48% of first-half 2026 revenue, or about $7.8 billion, flowed through Bedrock, Vertex, and Azure Marketplace. Those marketplace fees sit in sales, marketing, and partnerships, not cost of revenue, which helps explain how Anthropic can post an 85% gross margin in the first half after a much messier path in earlier periods.
The balance sheet is just as striking. Anthropic has raised $141.1 billion in aggregate and holds $112.4 billion in cash plus $19.3 billion in short-term investments, while implied lifetime burn is around $9.7 billion. That mix has invited a lot of chest-thumping about efficiency. It also comes with serious compute commitments, including up to $120.5 billion for NVIDIA-powered compute through SpaceXAI under agreements that extend into 2029.
The company is pitching the future around what it calls “Virtual Collaborators,” AI systems that work alongside humans on multi-step professional tasks and carry the access rights, system connections, and context their organizations grant them. Anthropic says these systems could serve a population of knowledge workers that is estimated to exceed one billion in 2028. That is a big claim, but the filing is built to support big claims: AI as a collaborator, not a chatbot; a business with strong margins, not just a research lab; and a company that wants investors to believe the next phase is bigger than the last.
My take — AI-written commentary, not fact-checked reporting
This is what AI hype looks like when it has accountants attached. The market loves a clean story, but Anthropic’s filing is really a lesson in how the cloud giants get paid twice: once for compute, once for distribution. The real novelty isn’t the slogan about collaborators, it’s that the bill is now large enough to make the slogans expensive.
Read more about this at: Latent Space
Related stories
Anthropic Says Revenue Jumped 45% to $16.8 Billion in Third Quarter
The New York Times · 1 week ago ·
3