Anthropic Tops OpenAI With $965B Valuation

Anthropic announced a $65 billion Series H financing that valued the company at $965 billion, overtaking OpenAI’s most recent $852 billion valuation. The round closed on Thursday and included previously committed capital.

The financing was led by Altimeter Capital, Dragoneer, Greenoaks and Sequoia Capital. The package includes $15 billion of previously committed investments, of which $5 billion came from Amazon. The new round nearly tripled Anthropic’s valuation since February, when it stood near $380 billion.

Anthropic reported an annualized revenue run rate of $47 billion, up from about $30 billion earlier this year and roughly $10 billion a year ago. OpenAI reported an annualized run rate near $24 billion in March, with company officials saying more recent figures are modestly higher.

The company has concentrated on enterprise products and developer tools. Recent updates include Claude Opus 4.5 and Claude Code, which gained traction among software developers. In January, Anthropic launched Cowork, an agentic AI product designed to automate nontechnical workplace tasks. The company also released Claude Mythos Preview, a restricted-access cybersecurity model for finding software vulnerabilities.

Enterprise adoption metrics tracked in April showed Anthropic overtaking OpenAI among business users. Data from corporate billing and routing platforms indicated Anthropic’s share of businesses rose to 34.4 percent while OpenAI’s share fell to 32.3 percent. On a platform that lets customers switch between models, Anthropic also gained market share. Ara Kharazian, lead economist at Ramp’s economics lab, observed that dominant firms can be displaced quickly in this market and described Anthropic’s recent gains as an example.

OpenAI has shifted strategy in response to competition in developer tools. The company agreed in March to acquire an AI coding startup to bolster its offerings and has narrowed its project priorities toward enterprise and coding products. OpenAI’s CEO outlined plans to invest about $1.4 trillion over the next eight years in infrastructure projects, including a global data-center network intended to secure compute capacity.

Anthropic’s CEO Dario Amodei cautioned about uncertainty in long-term infrastructure demand, warning that large, multi-year data-center buildouts carry timing risks. The company has moved toward per-token pricing to align customer costs with actual usage and to limit runaway compute consumption.

Industry executives have raised concerns about rapid token consumption and potential gaming of usage metrics. Ali Ghodsi, chief executive of Databricks, warned that systems can be misused to inflate token counts. OpenAI officials have acknowledged that unlimited or flat-rate pricing may be unsustainable as agentic workflows increase compute needs; its head of ChatGPT has indicated a need to reconsider flat-rate plans.

Both firms face operational challenges. Anthropic has spent to secure additional compute capacity and rented capacity from SpaceX after outages during periods of high demand. The company is also in discussions with U.S. defense authorities over national-security concerns tied to its systems. OpenAI faces pressure to justify large infrastructure spending through faster adoption and monetization.

Investors are weighing revenue growth, enterprise adoption, pricing models and infrastructure plans as the companies prepare for potential public listings. The new valuation places Anthropic ahead of OpenAI on paper, while both companies continue to expand product lines and pursue enterprise contracts.