OpenAI calls out to investors: computational power has now become a "hard constraint" for products, and we are running faster and harder than Anthropic.

date
07:48 10/04/2026
avatar
GMT Eight
This week, OpenAI revealed to its investors that, thanks to early aggressive investment in computing resources, the company has gained a crucial advantage in its competition with rival Anthropic PBC. Currently, Anthropic is considering an initial public offering (IPO).
This week, OpenAI revealed to its investors that, thanks to its early aggressive investment in computing resources, the company has gained a key advantage in the competition with rival Anthropic PBC. Currently, Anthropic is considering an initial public offering (IPO). According to a memo sent to some investors, the developer of ChatGPT stated that by "rapid and sustained" increases in computing capacity to support wider adoption of its software, OpenAI has surpassed Anthropic. This memo was released after Anthropic launched a more powerful AI model called Mythos. OpenAI pointed out in the memo that despite criticism of high costs, its ambitious infrastructure development allows it to better keep up with the growing demand for AI products. OpenAI wrote in the memo, "This difference is crucial because computational power has now become a limiting factor." OpenAI declined to comment. Anthropic, founded by former OpenAI employees, has seen increased demand in recent months due to the launch of new products and increased public support in its standoff with the Pentagon, sometimes struggling to keep its services online. In the memo to investors, OpenAI cited a report by Ben Thompson, author of the Stratechery blog, which indicated that computational limitations may have influenced Anthropic's decision to only release Mythos to select partners. OpenAI stated that they have 1.9 gigawatts of available computing capacity in 2025, three times more than the previous year (1 gigawatt can provide power for about 750,000 U.S. households). The company expects this number to increase to "low double digits" gigawatts next year and reach about 30 gigawatts by 2030. In comparison, OpenAI estimates that Anthropic will have 1.4 gigawatts of capacity by the end of 2025, and will have 7 to 8 gigawatts next year. OpenAI stated in the memo, "Even at the upper end of Anthropic's predicted range, our growth rate is significantly ahead and the gap is widening." In recent months, Anthropic has increased its investments in physical infrastructure for AI services, including a commitment to invest $50 billion in building data centers in the U.S. Additionally, Anthropic has expanded its strategic partnerships with Broadcom Inc. and Alphabet Inc.'s Alphabet Inc. Class C, to access around 3.5 gigawatts of computing power starting from 2027. Anthropic also has a diversified supplier base, collaborating with the three major cloud service providers, Alphabet Inc. Class C, Microsoft Corporation, and Amazon.com, Inc. In response to a request for comment, Anthropic cited a statement made by its Chief Financial Officer, Krishna Rao, regarding transactions with Broadcom Inc. and Alphabet Inc. Class C. Rao said, "This groundbreaking partnership with Alphabet Inc. Class C and Broadcom Inc. continues our cautious approach as we expand our infrastructure. We are making our most significant computational commitment to date to keep up with this unprecedented growth." In the past, Anthropic's spending strategy has typically been more conservative than OpenAI's. OpenAI plans to invest around $600 billion by 2030 in data centers and chips, and recently raised $122 billion to support these commitments. However, considering OpenAI is not expected to be profitable in the coming years, the speed of its spending is still being questioned. On Thursday, OpenAI announced it will suspend a infrastructure project in the UK due to energy cost issues. Dario Amodei, CEO of Anthropic, mentioned the company's spending plans at the end of last year, saying, "As a company, we will manage responsibly as much as possible. I think some players are being 'YOLO.'" In the investor memo, OpenAI described Amodei's views as underestimating the level of demand for more AI products in the market. The company stated, "In hindsight, this caution may have been more of an underestimation of the speed at which demand would arrive rather than discipline."