OpenAI's Chief Financial Officer Sarah Friar stated in a blog post last Sunday that the company's annualized revenue for 2025 has surpassed $20 billion, with revenue growth being closely tied to the expansion of its computing power scale. She also indicated that the company is positioning 2026 as the year of "practical application." Friar mentioned that OpenAI's annualized revenue for 2025 exceeded $20 billion, a significant increase from the $6 billion figure in 2024 and a tenfold jump from the $2 billion recorded in 2023. She revealed that OpenAI's computing power scale surged from 0.6 gigawatts in 2024 to 1.9 gigawatts in 2025. She further pointed out that the company's weekly active user and daily active user metrics continue to hit new all-time highs, "representing 'unprecedented growth at scale.' Moreover, we firmly believe that investing more compute resources during these periods will accelerate customer product adoption and monetization." Simultaneously, Friar stated, "The urgent priority is to bridge the gap between what AI can currently achieve and how individuals, businesses, and nations use AI in their daily lives. The opportunity is immense and pressing, particularly in healthcare, science, and enterprise, because better intelligence directly translates into better outcomes." Friar elaborated on how OpenAI views the commercialization strategy for its services, such as ChatGPT, while ensuring the necessary computational capacity for these products. She emphasized that OpenAI's revenue is directly correlated with the availability of its technological infrastructure. Friar highlighted that OpenAI's computational capacity grew from 0.2 gigawatts in 2023 to approximately 1.9 gigawatts in 2025, while the company's annual revenue correspondingly surged from $2 billion in 2023 to over $20 billion last year. Friar's blog post comes at a time when the focus on artificial intelligence by OpenAI and the tech industry faces scrutiny, given the massive investments required for building data centers and securing the energy and components for this cutting-edge technology, which has so far yielded little revenue return for investors. Among these significant deals is an agreement reached between OpenAI and NVIDIA last September. Under this agreement, NVIDIA indicated it would commit $100 billion to support the AI startup, aiding in the construction and deployment of at least 10 gigawatts of NVIDIA systems. A "gigawatt" is a unit of power measurement, with 10 gigawatts being roughly equivalent to the annual electricity consumption of 8 million U.S. households. However, in November last year, NVIDIA informed investors that its agreement with OpenAI came with "no guarantee" of progressing smoothly from the announcement phase to a formal contract stage. Friar wrote, "Achieving world-class compute capabilities requires commitments made years in advance, and the development path is not always smooth." She also noted that this system demands discipline. Friar also mentioned that three years ago, OpenAI relied on just a single compute supplier, but it has now established partnerships with multiple suppliers, creating a diversified ecosystem. She stated, "In a market where 'who can scale depends on who can get compute,' we are able to plan, finance, and deploy capabilities with confidence." Friar expressed that OpenAI's business model should align with and evolve alongside the services it provides. She wrote, "As intelligence integrates into scientific research, drug discovery, energy systems, and financial modeling, new economic models will emerge." It is worth noting that Friar's blog was published after OpenAI stated last week its plans to begin testing ads for some ChatGPT users in the U.S., as the company prepares for a potential public listing later this year. Friar said, "Monetization should blend naturally with the user experience. If it doesn't add value, it shouldn't exist."
Comments