NextFin News - OpenAI Chief Financial Officer Sarah Friar announced in a blog post on Sunday that the company’s annualized revenue has surpassed $20 billion in 2025, marking a staggering increase from the $6 billion reported in 2024. According to Reuters, this financial surge is closely aligned with a dramatic expansion in the company’s infrastructure, with computing capacity rising to 1.9 gigawatts (GW) from 0.6 GW a year earlier. Friar noted that the Microsoft-backed firm continues to see all-time highs in both weekly and daily active users, signaling sustained demand for its generative AI services.
The revenue milestone comes as OpenAI aggressively diversifies its monetization strategies to offset the high costs of model development. Last week, the company began testing advertisements within ChatGPT for select users in the United States. Furthermore, OpenAI is preparing to move beyond software; Chris Lehane, the company’s policy chief, confirmed that the firm is on track to unveil its first physical device in the second half of 2026. For the current year, Friar emphasized that the company will prioritize "practical adoption" in specialized sectors such as healthcare, science, and enterprise automation.
The correlation between OpenAI’s revenue and its 1.9 GW computing capacity highlights the "compute-as-currency" reality of the modern AI industry. By tripling its power intake, OpenAI has effectively expanded the throughput of its Large Language Models (LLMs), allowing for more concurrent users and more complex API calls. This 216% increase in power capacity serves as the physical foundation for the 233% increase in annualized revenue, suggesting that for the time being, AI growth remains a linear function of hardware scaling. However, this dependency also underscores the company's vulnerability to energy costs and chip supply chains.
To mitigate the risks associated with such heavy infrastructure requirements, Friar detailed a strategy of maintaining a "light" balance sheet. Rather than pursuing a capital-intensive model of owning data centers, OpenAI is utilizing flexible contracts and partnerships with cloud providers. This approach allows the company to remain agile, pivoting between different hardware types and providers as the technological landscape shifts. This financial flexibility is critical as the company transitions from simple chat interfaces to "agents"—autonomous systems capable of executing multi-step workflows and taking actions across various software tools.
The introduction of advertising and the move into hardware represent a strategic pivot toward sustainability. While the $20 billion revenue figure is impressive, the cost of maintaining 1.9 GW of capacity and the research and development for next-generation models like GPT-5 likely keeps the company’s net margins thin or negative. By entering the hardware market in 2026, OpenAI aims to create a proprietary ecosystem, reducing its reliance on third-party platforms and potentially capturing higher-margin consumer data and subscription loyalty. This move mirrors the vertical integration strategies of tech giants like Apple, suggesting that OpenAI no longer views itself as merely a research lab, but as a full-stack technology conglomerate.
Looking forward, the focus on "practical adoption" in health and science indicates a shift toward high-value, low-churn enterprise markets. As U.S. President Trump’s administration emphasizes American leadership in emerging technologies, OpenAI’s rapid scaling positions it as a national champion in the global AI race. However, the transition to an ad-supported model and the launch of physical devices will test the brand’s relationship with its user base. The coming year will determine if OpenAI can convert its massive computing footprint into a profitable, self-sustaining business model that can withstand the intensifying competition from well-capitalized rivals like Google and Meta.
Explore more exclusive insights at nextfin.ai.