Anthropic announced in early April 2026 that its annualized revenue run rate has crossed $30 billion, surpassing OpenAI's reported $25 billion ARR. The milestone is striking because Anthropic was founded only in 2021 by former OpenAI employees who left citing safety concerns, and for years was considered the underdog to OpenAI's dominant market position. The company simultaneously announced an expanded compute partnership with Google and Broadcom that will deliver approximately 3.5 gigawatts of next-generation TPU capacity starting in 2027.
The revenue crossover has multiple explanations. Claude Opus 4.7 has won enterprise contracts from companies that prefer Anthropic's transparency reports and safety commitments. Anthropic's API is increasingly embedded in software products, coding tools, and customer service platforms. And the company's Constitutional AI approach — which trains models to follow a set of stated principles — appears to have become a selling point for risk-conscious buyers in regulated industries like healthcare, finance, and law.
OpenAI, meanwhile, is not standing still. The company is reportedly taking early steps toward a public offering, potentially as soon as late 2026, and has diversified into hardware partnerships, enterprise software, and consumer apps. Both companies are growing fast enough that a few months of relative position could flip again. The real story is that the AI services market has become large enough to support multiple $25B-plus businesses simultaneously — something almost no analyst predicted three years ago.
For students, the competitive dynamics here offer a lesson in how markets reward different strategies at different stages. OpenAI's first-mover advantage gave it the consumer mindshare and developer ecosystem. Anthropic's differentiated positioning on safety gave it a wedge in enterprise sales. Neither approach is universally right; understanding both is part of understanding how the AI industry actually works, as opposed to how it is portrayed in headlines.