Anthropic Reportedly Targets $2 Trillion IPO Valuation as Compute Deals and Revenue Accelerate

SAN FRANCISCO — Anthropic, the AI safety company behind the Claude family of models, has filed confidentially for an initial public offering and investors are targeting a valuation of approximately $2 trillion for a debut expected as soon as October 2026, according to multiple reports citing unnamed sources familiar with the matter. If the offering prices near that figure, it would be the largest IPO in history by a wide margin. Anthropic's financial trajectory has been extraordinary. Revenue run-rate scaled from roughly $30 billion in April 2026 to $47 billion in May — the month it closed a $65 billion Series H funding round — and to an estimated $69-74 billion by late July, according to third-party estimates. The company reported positive adjusted operating income of $559 million in the second quarter of 2026, its first quarterly operating profit, versus $787 million in revenue in the same quarter of 2025. Annualized revenue run-rate crossed $47 billion in May. Anthropic's revenue comes primarily from API access to Claude models (per-token billing for developers and enterprises), Claude subscriptions (Pro/Team/Enterprise) and large cloud partner agreements with Amazon Web Services and Google Cloud, both of which are also investors. The company's enterprise-first positioning — pitching Claude as the safest and most steerable frontier model — has made it the default choice for regulated industries and coding workloads. On the compute side, Anthropic signed a $10 billion, six-year agreement with Volta Infra Holdings Ltd., a cloud startup backed by Nvidia, for capacity at a Volta-managed data center in Norway partnered with Bitdeer Technologies Group. The deal follows Anthropic's reported acquisition of Decart, an efficiency startup, for approximately $6 billion, and a $9.1 billion, 20-year compute lease with Riot Platforms at its Rockdale, Texas campus. Morgan Stanley is providing $573 million in interim financing for the Riot deal. Anthropic has also been publicly warning about risks from AI self-improvement trajectories, and has indicated it is considering pausing certain development activities. The unusual admission from a frontier AI developer has added urgency to policy discussions in Europe and the United States about how to govern the most capable AI systems.