← Back to feed

OpenAI closed a $122 billion funding round on March 31, valuing the company at $852 billion — the largest private fundraise in history, exceeding the GDP of more than 130 countries and surpassing Meta’s 2021 peak valuation.

Key Numbers

• $122B total committed capital (up from $110B announced in February)
• $852B post-money valuation — on par with Berkshire Hathaway
• $2B/month in current revenue, up from $1B/quarter at end of 2024
• $13.1B total revenue in 2025
• $64B+ total capital raised since 2015 founding
• Breakeven projected no earlier than 2030 — not yet profitable

Who Invested

The round was co-led by SoftBank, Andreessen Horowitz, and D.E. Shaw Ventures:

• Amazon: $50B (largest single commitment) — $15B immediate, $35B contingent on IPO or AGI milestone
• Nvidia: $30B
• SoftBank: $30B
• Additional: Microsoft, MGX, TPG, T. Rowe Price, Blackstone, Sequoia, Temasek, Thrive Capital, ARK Invest, and others
• $3B+ raised from individual investors through bank channels — first time OpenAI has done this

OpenAI also expanded its revolving credit facility to $4.7B, backed by JPMorgan, Goldman Sachs, Morgan Stanley, and others. The facility was undrawn at close.

The Infrastructure Bet

The capital allocation is not primarily about model development — it’s physical infrastructure. OpenAI has committed to a $300B, five-year cloud partnership with Oracle delivering 4.5 gigawatts of data center capacity by 2027. That’s purpose-built AI infrastructure engineered to OpenAI’s thermal and power density requirements, at a scale beyond conventional enterprise projects.

Amazon’s $35B contingent tranche is tied to OpenAI deploying enterprise tools via AWS — a plan reportedly being challenged internally by Microsoft, which remains a major backer and Azure partner.

What It Means

At $852B, OpenAI is the most valuable private company in history. Enterprise now accounts for more than 40% of revenue and is on track to match consumer revenue by end of 2026. ChatGPT has 900M+ weekly active users.

The round cements the infrastructure thesis: the next phase of AI competition is not about model benchmarks but gigawatt-scale compute access. OpenAI needs to fund training runs that require more power than mid-sized cities. The question now is whether the revenue trajectory — currently doubling roughly every 12 months — can catch the capital commitment before 2030.