Facing ongoing doubts about whether its massive investments can yield equivalent returns, OpenAI's Chief Financial Officer, Sarah Fleur, recently provided a strong response through official channels. Over the past three years, OpenAI has not only made remarkable technological advancements but also achieved an astonishing "tenfold" leap in its commercialization capabilities and infrastructure scale.

According to data, OpenAI's computing power resources have surged from 0.2GW in 2023 to 1.9GW in 2025, representing a cumulative increase of about 9.5 times. Excitingly, the company's revenue growth has closely followed the expansion of its computing power: starting at $2 billion in 2023, it is expected to exceed $20 billion in 2025. Fleur emphasized that this growth curve proves the rationality of large-scale infrastructure investment, and the increase in revenue continues to provide funding for the next stage of intelligent evolution.

To reduce reliance on a single supplier, OpenAI has transitioned from initially relying entirely on Microsoft to a multi-supplier model, entering into agreements worth hundreds of billions of dollars with giants such as NVIDIA, AMD, and Oracle. In terms of operational strategy, the company adopts "precise computing power allocation," using top-tier hardware to train cutting-edge models, while employing low-cost facilities for high-concurrency scenarios like large-scale inference, thereby achieving a balance between efficiency and expenditure. Currently, OpenAI has established a diversified revenue matrix consisting of ChatGPT subscriptions, API services, and advertising e-commerce, steadily moving toward becoming an essential foundation of the global economy.

Key Points:

  • 📈 Revenue and Computing Power Both Tenfold: OpenAI has achieved approximately tenfold growth in both revenue and computing power over the past three years, with projected revenue exceeding $20 billion in 2025.

  • 🤝 Diversified Supply Matrix: The company has broken away from a single-supplier model, signing agreements worth hundreds of billions of dollars with major companies such as Microsoft and NVIDIA, enhancing control over computing power deployment.

  • 💰 Diversified Business Model: The company generates revenue through multiple channels such as subscriptions, API services, and advertising, and plans to explore new models like licensing and outcome-based pricing.