OpenAI's Staggering Financial Blueprint: $600 Billion in Compute, $100 Billion in Losses, and a Bet on the Future of Intelligence
OpenAI has laid out what may be the most ambitious — and expensive — business plan in technology history. According to financial projections shared with investors last week, the company expects to spend approximately $600 billion on computing infrastructure by 2030, generate more than $280 billion in cumulative revenue, and absorb over $100 billion in losses along the way.
The numbers, first reported by CNBC and Bloomberg, represent a significant recalibration from CEO Sam Altman's earlier statements about $1.4 trillion in infrastructure commitments. The revised figures offer investors a more defined timeline and a spending plan that ties more directly to anticipated revenue growth — a concession, perhaps, to mounting concerns that the company's ambitions have outpaced its ability to generate returns.
To finance this expansion, OpenAI is reportedly assembling the largest private capital raise in history: a funding round that could exceed $100 billion, valuing the company between $730 billion and $850 billion. Nvidia is in talks to invest up to $30 billion, with SoftBank, Amazon, and other strategic investors expected to contribute the lion's share — roughly 90 percent of the total.
The revenue picture, at least on paper, is promising. OpenAI generated $13.1 billion in revenue in 2025, surpassing its own $10 billion target, while spending $8 billion against a $9 billion budget. ChatGPT now claims between 800 and 900 million weekly active users worldwide, and the company expects roughly equal contributions from its consumer and enterprise businesses going forward. It has also begun testing advertising and commerce features within ChatGPT, including sponsored product placements and shopping integrations — a notable departure from its original mission-driven ethos.
But the sheer scale of what OpenAI is attempting to build sets it apart from any software company that has come before it. The $600 billion compute investment encompasses hyperscale data centres, advanced AI chips, high-speed networking infrastructure, and the energy systems required to power it all. The company scaled from approximately 0.6 gigawatts of computing capacity in 2024 to 1.9 gigawatts in 2025, and the trajectory only steepens from here through the Stargate initiative, a joint infrastructure venture that aims to establish a new class of AI-dedicated facilities.
The economics of frontier AI remain punishing. Training and operating the most capable models requires continuous, enormous expenditure on semiconductor procurement, research and development, and data centre operations. High-performance GPUs remain expensive and supply-constrained, which is why OpenAI has invested in custom silicon to reduce its long-term dependence on third-party chip suppliers.
Meanwhile, the company continues to push the boundaries of what its models can do. Earlier this month, GPT-5.2 made headlines by independently deriving and proving a new formula in theoretical physics — what OpenAI called AI's first original contribution to the field. Last week also saw reports that the company has launched a 200-person effort to develop AI-powered consumer devices, including a smart speaker, smart glasses, and a smart lamp, with the earliest hardware expected no sooner than February 2027.
The question that hovers over all of this is whether the economics will ever work. OpenAI is essentially asking investors to fund the construction of a new kind of global infrastructure — one built not on roads or fibre optic cables, but on the raw computational power needed to run increasingly intelligent machines. The company is projecting that it will remain unprofitable for much of the decade, a timeline that would test the patience of even the most committed backers.
Yet the scale of investment flowing into the round suggests that the world's largest technology companies and sovereign wealth funds have decided the risk is worth taking. If OpenAI is right about the trajectory of artificial intelligence, the $600 billion it plans to spend could look like a bargain in hindsight. If it is wrong, it will stand as one of the most spectacular financial miscalculations in corporate history.
Either way, the AI industry has entered a phase where the numbers have moved beyond what most people can intuitively grasp — and OpenAI is leading the charge into that uncharted territory.
