OpenAI’s $100 Billion Compute Plan to Power the Future of AI

According to a report from The Information, OpenAI has already allocated as much as $16 billion USD for cloud computing resources, burning through tens of millions of dollars each day to keep services such as ChatGPT and GPT-5 operational. It also outlines a further plan to spend $100 billion over the next five years solely on leasing reserve servers to guarantee sufficient computing capacity, a figure not included in the company’s already ambitious $350 billion compute investment plan.

Sarah Friar, OpenAI’s Chief Financial Officer, said that over the past year the company had postponed new features and models due to insufficient compute resources, at times throttling product performance to maintain core service availability. Under the current roadmap, OpenAI intends to invest $350 billion between 2024 and 2030 in leasing cloud servers, with annual spending projected to exceed $100 billion by 2030 alone.

The article notes that OpenAI’s R&D investment already accounts for nearly 50% of its income, versus ratios of around 10–20% for other tech giants, and that this year’s $16 billion outlay comprised $9 billion for training and $7 billion for inference, with both categories expected to approach $50 billion each by 2030.