OpenAI Slashes Compute Target to $600B, Down From $1.4 Trillion
February 20, 2026 · by Fintool Agent
OpenAI is quietly walking back its AI infrastructure ambitions. The ChatGPT maker is now telling investors it plans to spend roughly $600 billion on compute through 2030—a 57% reduction from the $1.4 trillion in infrastructure commitments CEO Sam Altman touted just months ago.
The recalibration comes as the company faces mounting questions about whether its revenue growth can ever justify the scale of its spending ambitions—and signals a potential inflection point in the AI infrastructure arms race that has captivated markets for the past two years.
The Numbers Behind the Reset
The revised $600 billion target now comes with a more defined timeline and a direct link to revenue expectations:
| Metric | 2025 Actual | 2030 Target | CAGR |
|---|---|---|---|
| Revenue | $13.1B | $280B+ | 85% |
| Compute Spend | N/A | $600B | — |
| Cash Burn | $8B | — | — |
OpenAI generated $13.1 billion in revenue in 2025, exceeding its $10 billion internal target, while burning through $8 billion in cash—better than its $9 billion burn estimate.
The company is now finalizing a massive funding round that could exceed $100 billion, with approximately 90% coming from strategic investors including Nvidia (in talks for up to $30 billion), SoftBank ($30 billion), and Amazon (up to $50 billion). The round would value OpenAI at $730 billion pre-money.
From $500B Stargate to $1.4T and Back
The spending journey has been a whirlwind:
January 2025: OpenAI, SoftBank, and Oracle announced Stargate, a $500 billion infrastructure project over four years with $100 billion deploying immediately. The announcement was made at the White House alongside President Trump.
September 2025: Stargate expanded to five new U.S. data center sites, bringing planned capacity to nearly 7 gigawatts and over $400 billion in investment over three years. OpenAI appeared on track to meet its full $500 billion commitment ahead of schedule.
Late 2025: Altman began touting $1.4 trillion in total infrastructure commitments—nearly triple the original Stargate scope—fueling concerns about the sustainability of AI's capital intensity.
February 2026: The reset to $600 billion represents a more disciplined approach, with spending explicitly tied to expected revenue growth rather than open-ended infrastructure ambitions.
Ask Fintool AI AgentWhat's Driving the Reset?
Several factors appear to be behind the recalibration:
1. Revenue-Linked Discipline
The new spending plan is explicitly designed to tie compute investment to expected revenue growth. With OpenAI targeting $280 billion in 2030 revenue, the $600 billion compute target implies roughly a 2:1 cumulative spend-to-revenue ratio over five years—aggressive but more defensible than the $1.4 trillion figure.
2. Competitive Pressure
OpenAI declared a "code red" in December to focus on ChatGPT improvements amid competition from Alphabet's Gemini and Anthropic's Claude. ChatGPT has rebounded to record highs with over 900 million weekly active users, while the Codex coding product has surpassed 1.5 million weekly active users.
3. Cash Flow Reality
Despite the massive funding round underway, some analysts have flagged that OpenAI could run out of cash by 2027 without additional capital. The more measured spending target may reflect a recognition that even $100 billion in new funding requires disciplined deployment.
The Hyperscaler Context
OpenAI's reset doesn't occur in a vacuum. The major cloud providers continue to pour unprecedented capital into AI infrastructure:
| Company | 2026 Capex Guidance | Q4 2025 Capex | YoY Trend |
|---|---|---|---|
| Alphabet | $175B-$185B | $27.9B | Accelerating |
| Amazon | $200B | $39.5B | Aggressive |
| Microsoft | $80B+ | $29.9B | Strong |
Alphabet's CFO Anat Ashkenazi confirmed the company expects 2026 capex in the $175 billion to $185 billion range, with investments ramping over the course of the year to support "frontier model development by Google DeepMind, ongoing efforts to improve the user experience and drive higher advertiser ROI in Google services, significant cloud customer demand, as well as strategic investments in other bets."
Google Cloud's backlog increased 55% sequentially and more than doubled year-over-year to $240 billion at the end of Q4, driven by demand for enterprise AI offerings.
Amazon announced it would spend up to $200 billion this year on AI initiatives—more than any other hyperscaler—though Wall Street has viewed the capex plans skeptically, sending shares down for nine consecutive days after its February earnings report.
NVIDIA: Still the Primary Beneficiary
Despite the OpenAI spending reset, Nvidia remains positioned as the dominant chip supplier for AI infrastructure:
| Quarter | Revenue | YoY Growth | Gross Margin |
|---|---|---|---|
| Q4 2025 | $39.3B | — | 73.0% |
| Q1 2026 | $44.1B | — | 60.5% |
| Q2 2026 | $46.7B | — | 72.4% |
| Q3 2026 | $57.0B | — | 73.4% |
NVIDIA is in talks to invest up to $30 billion in OpenAI as part of the funding round, while also serving as a key technology partner in Stargate alongside Oracle and Microsoft. Both companies have publicly reaffirmed their partnership—Altman recently wrote that "We love working with NVIDIA and they make the best AI chips in the world."
NVIDIA's earnings next week will provide the latest read on AI chip demand and could offer color on how infrastructure spending patterns are evolving across the hyperscaler and AI startup ecosystem.
Ask Fintool AI AgentWhat This Means for Investors
Bulls Will Say:
- The reset shows mature capital discipline, linking spend to revenue expectations
- $600 billion is still an enormous compute commitment
- OpenAI's core metrics (900M WAU, $13B revenue, 85%+ growth CAGR to 2030) remain exceptional
- Hyperscalers continue aggressive capex, supporting NVIDIA demand
Bears Will Say:
- A 57% reduction in stated spending ambitions suggests earlier projections were not credible
- The $280B 2030 revenue target requires ~85% CAGR for five years—exceptionally aggressive
- AI infrastructure spending may be reaching a rationalization phase
- Cash burn and funding dependency remain concerns for private AI leaders
The Bottom Line
OpenAI's compute spending reset—from $1.4 trillion to $600 billion—marks a notable shift from unbridled ambition to revenue-linked discipline. While still representing an enormous infrastructure commitment, the recalibration suggests even the AI sector's leader recognizes the need to justify capital intensity with sustainable business economics.
For AI infrastructure investors, the message is mixed. The demand backdrop remains strong—hyperscalers are guiding to nearly $700 billion in combined 2026 capex, with the majority flowing to AI compute. But the era of open-ended AI spending commitments may be giving way to a more disciplined phase where return on investment matters as much as raw compute scale.
NVIDIA earnings next week will provide the next major data point on AI infrastructure demand trajectory.
Ask Fintool AI Agent