Anthropic revenue surge fuels trillion-dollar valuation talk

Anthropic’s annualised revenue reportedly topped $30 billion earlier this month, a jump that would place the Claude maker among the fastest-growing software businesses on record. The figure, reported by The Information and relayed by The Decoder, has triggered investor talk of valuations approaching $1 trillion—numbers usually reserved for the largest public tech platforms.

The reporting describes a growth curve driven less by consumer chatbots than by enterprise contracts and developer tooling. A key product is Claude Code, Anthropic’s coding assistant, alongside a broader push into “coworker” style agents that can take on multi-step tasks. That matters because it shifts AI from a novelty interface to a budget line inside companies: once an assistant is embedded in engineering workflows, usage becomes recurring and hard to unwind.

The same sources say Anthropic’s economics have improved sharply. Gross margins are described as swinging from deeply negative in 2024 to roughly 40% the following year, implying either cheaper inference, better pricing power, or customers willing to pay for higher-end models. In practical terms, that is the difference between a company that must constantly raise capital to subsidise compute and one that can dictate terms to investors.

That leverage shows up in the financing details. The Information is reported to have told investors they were expected to write cheques in the hundreds of millions for Anthropic’s September 2025 round, and that the company has been able to resist pressure for additional rounds in the near term. The Decoder reports that CFO Krishna Rao’s team has already fielded offers valuing the company at about $800 billion.

The story also underlines how concentrated the AI economy is becoming. Training and serving frontier models requires access to scarce inputs—chips, power, datacentre capacity, and distribution partnerships—so the firms that reach scale first can widen the gap by reinvesting cash flow into more compute. For customers, the appeal is straightforward: buying tokens is easier than hiring teams, and the cost can be booked as operating spend rather than long-term headcount.

For everyone else, the arithmetic is less forgiving. A trillion-dollar private valuation would not just be a paper milestone; it would signal that the market expects a small number of model providers to sit between businesses and their software stack, charging rent on everyday work. The rest of the AI ecosystem—from open-source projects to smaller tooling startups—then competes in the shadow of companies that can price below cost to win adoption and raise later on better terms.

Anthropic has not publicly confirmed the revenue figure, but the numbers now circulating are already shaping expectations. In this market, the valuation is often decided before the audited accounts arrive.