OpenAI is rapidly converting technological advantages into commercial momentum. On Sunday, the company's Chief Financial Officer, Sarah Friar, publicly disclosed that **OpenAI's annualized revenue for 2025 has exceeded $20 billion, representing an explosive growth of over 230% compared to $6 billion in 2024**. This surge is not accidental but closely tied to the rapid expansion of its computing infrastructure—by 2025, OpenAI's total computing power had reached 1.9 gigawatts (GW), more than doubling from 0.6 gigawatts in 2024.

Friar emphasized that computing power has become the core engine driving OpenAI's revenue growth. "Every additional watt of computing power means more users, more API calls, and more enterprise contracts." With full support from Microsoft Azure cloud platform, OpenAI not only ensures stable training and inference for the GPT series models but also supports its product portfolio—including ChatGPT, Operator agents, and enterprise API services—to achieve high concurrency and low latency global delivery.

User growth data is equally impressive. Thanks to performance improvements and expanded scenarios, **OpenAI's weekly and daily active user numbers continue to break historical records**. Whether it's individual users using ChatGPT to handle daily tasks, developers calling APIs to build AI applications, or enterprises deploying customized agents, strong demand is driving a sharp upward trend in revenue.

Notably, the $20 billion annualized revenue places OpenAI among the most profitable AI companies globally, even surpassing some emerging business segments of traditional tech giants. The 1.9 gigawatt computing power is equivalent to the electricity consumption of a medium-sized city, highlighting its infrastructure characteristics as a "utility" provider in the AI era.

Amid the intense competition for AGI, OpenAI is accelerating its lead through the "computing power—users—revenue" flywheel effect. While competitors are still struggling with thousand-card clusters, OpenAI has quietly turned computing power into real money and built an unreplicable ecological moat. This AI revolution driven by electricity has only just reached its climax.