According to the latest report from CNBC, OpenAI directly mocked its long-time rival Anthropic in a memo addressed to investors. OpenAI believes that the latter is significantly lagging in computing power and facing serious resource constraints, with the gap between them continuing to widen.

The Computing Power Gap Continues to Widen

OpenAI revealed an ambitious expansion plan in the memo, aiming to increase its computing power to 30 gigawatts by 2030. By comparison, OpenAI predicts that Anthropic's computing power will only reach 7 to 8 gigawatts by the end of 2027. Even at the highest estimate, its growth rate is far behind OpenAI's.

This gap stems from different strategic choices between the two companies. OpenAI has opted for an aggressive infrastructure expansion model, leveraging the "compounding advantage" to reduce costs and enhance model intelligence. In contrast, Anthropic is seen as taking a more conservative investment path, which puts it at a disadvantage in the computing power competition.

The $1 Trillion IPO Race

Currently, the combined valuation of these two AI giants exceeds $1 trillion, and both are actively pushing forward their listing plans. OpenAI emphasizes that stronger infrastructure leads to higher cost-effectiveness per unit of intelligence, a leading advantage that allows it to provide free tools to hundreds of millions of users while generating higher revenue through premium products.

At the same time, Anthropic, founded by former OpenAI executives, is also focusing on the enterprise market, recently launching a new model and a cybersecurity project. However, in OpenAI's view, without sufficient computing power and resources, the competitor will find it difficult to maintain long-term commercial competitiveness amid the pressure from industry giants like Google and Meta.