Market news: OpenAI’s annualized revenue approaches US$70 billion, and enterprise revenue doubles in three months

Financial Associated Press, September 29 (Editor Shi Zhengcheng)On Tuesday evening, Beijing time, some media quoted people familiar with the matter as reporting that, OpenAI approaches $70 billion in annual recurring revenue (ARR), up more than 70% since the beginning of the third quarter.

其中， Enterprise business revenue has increased by more than 100% since July, and the new revenue from the consumer side in the third quarter has also exceeded the annual increase in 2025.

It should be noted that the annualized revenue run rate converts recent revenue levels into one year and is used to measure the company’s current revenue capabilities.

Looking back at key nodes in the past few years, OpenAI’s revenue growth has experienced a rapid jump from billions of dollars to tens of billions of dollars.

According to data disclosed by OpenAI Chief Financial Officer Sarah Fryer on the official website earlier this year, the company’s annual revenue will be approximately US$2 billion in 2023, rising to US$6 billion in 2024, and further exceeding US$20 billion in 2025.

Entering 2026, this growth curve continues to rise. In March this year, it was reported that OpenAI’s annualized revenue as of the end of February had exceeded US$25 billion, an increase of approximately 17% from US$21.4 billion at the end of 2025. By July, the company’s annual revenue run rate further exceeded US$40 billion.。

In horizontal comparison, OpenAI’s competitor Anthropic also continues to grow rapidly.

According to official data released by Anthropic in April this year, the company’s annualized revenue at the end of 2025 will be approximately US$9 billion, which is significantly lower than OpenAI’s level in the same period. By the time a new round of financing was announced on February 12, 2026, Anthropic’s annual revenue had reached US$14 billion; on April 6, the company announced that this indicator exceeded US$30 billion.

Enterprise customers and programming tools are important forces supporting this round of expansion. As of the end of July, Anthropic’s annualized revenue run rate exceeded $65 billion, which is higher than OpenAI’s level of more than $40 billion at the time.

On September 18, there were reports that Anthropic expected annual revenue to exceed US$100 billion this year.

It should be noted that there are obvious differences in the revenue recognition methods of the two companies.

Previously, media reports quoted people familiar with the matter as saying that when selling model services through cloud partners, OpenAI only recognized its own share of the transaction as revenue for some transactions, while Anthropic included the entire amount paid by customers as revenue.

From a capital and industry perspective, to support the entire industry’s huge hardware investment, AI commercialization still needs to continue to accelerate.

Bain pointed out that four types of businesses are expected to contribute an additional $4.2 trillion in revenue needed to support the construction of AI infrastructure.

First, cutting-edge model developers are replacing search engines and opening up new revenue streams through product placement. Second is the rise of “autonomy of everything,” including self-driving cars, trucks, drones, and other forms of industrial automation. The third is physical AI, including simulation, digital twins and robotics. Finally, there is still about a $3 trillion revenue gap that needs to be filled through the development of new products.

This also means that for leading manufacturers such as OpenAI, the current rapid growth is just the starting point: for chip procurement and data center expansion to continue, the entire industry must still convert the increasing computing power into a wider range of payment needs, and ultimately realize it into actual revenue and return on investment.