OpenAI Slashes Infrastructure Spending Forecast to $600 Billion as IPO Looms

2026-02-21 companies

San Francisco, Saturday, 21 February 2026.
OpenAI recalibrated its 2030 infrastructure target to $600 billion, effectively halving previous estimates of $1.4 trillion, marking a pivotal shift in capital strategy as it prepares for a public listing.

Strategic Recalibration Ahead of Public Listing

In a decisive move to align financial expectations with operational reality, OpenAI has informed investors that it now targets approximately $600 billion in total compute spending through 2030 [1][7]. This updated guidance marks a sharp deviation of -57.143% from the $1.4 trillion infrastructure commitment previously touted by CEO Sam Altman, which had aimed to develop 30 gigawatts of computing resources [7][8]. The dramatic reduction in projected capital expenditure appears to be a calculated step to demonstrate fiscal discipline as the company lays the groundwork for a potential initial public offering (IPO) as early as the fourth quarter of 2026 [4][7].

Revenue Growth and Financial Health

Despite the slashed infrastructure budget, OpenAI’s revenue forecasts remain robust. The company projects its annual revenue will surpass $280 billion by 2030, a sum expected to be divided nearly equally between its consumer and enterprise business units [1][7]. This bullish long-term outlook is supported by strong recent performance; in 2025, OpenAI generated $13.1 billion in revenue, beating its internal projection of $10 billion [7]. Furthermore, the company demonstrated tighter cost control than anticipated, with a 2025 burn rate of $8 billion against a target of $9 billion [1][7]. However, the path forward remains capital-intensive, with reports indicating the company anticipates an additional $112 billion in cash burn through the end of the decade [3].

Historic Capital Injection

To finance its ambitious roadmap, OpenAI is finalizing a massive funding round seeking more than $100 billion, a move that could value the company at approximately $830 billion [4][7]. As of February 19, 2026, semiconductor giant Nvidia is in discussions to invest up to $30 billion, a deal that implies a pre-money valuation of $730 billion [1]. The round is expected to be dominated by strategic investors, with SoftBank reportedly slating $30 billion in investment across three installments, and Amazon potentially contributing up to $50 billion [4]. If completed, this capital raise would stand as one of the largest private funding events on record, providing OpenAI with a substantial war chest to navigate the competitive AI landscape [4][7].

Margin Pressures and Market Volatility

While the capital influx validates OpenAI’s market position, internal metrics highlight the rising costs of maintaining dominance. The expense of ‘inference’—the computing power required to run AI models for users—quadrupled in 2025, causing the company’s adjusted gross margin to contract to 33% from 40% in 2024 [7]. This margin compression coincides with a broader market reevaluation of AI economics. In early February 2026, skepticism regarding AI investment returns triggered a market selloff that wiped over $1 trillion in value from major tech companies [5]. With competitors like Google, Amazon, Meta, and Microsoft planning a combined $650 billion in capital spending for 2026 alone [5], OpenAI’s decision to moderate its 2030 spending target to $600 billion suggests a strategic pivot toward sustainable efficiency over unbridled expansion.

Sources

Artificial Intelligence Capital Expenditure