OpenAI $500B Valuation: $6.6B Sale Leaves $3B Unsold

💡 TL;DR - The 30 Seconds Version

💰 OpenAI closed a $6.6 billion secondary share sale at a $500 billion valuation on Thursday, making it the world's most valuable private company and surpassing SpaceX's $456 billion mark.

🤔 Employees left $3.4 billion in authorized shares unsold—only 64% participated despite the rich pricing—suggesting either confidence in higher future valuations or caution about ongoing corporate restructuring.

📉 The company burned $2.5 billion in the first half of 2025 against $4.3 billion in revenue, with profitability not expected until 2029 when it projects reaching $125 billion in annual revenue.

🎯 The tender serves as a retention weapon against Meta's aggressive recruiting, which has poached at least 10 OpenAI researchers with compensation packages exceeding $100 million since June 2025.

⚖️ OpenAI's conversion from nonprofit to public benefit corporation remains pending regulatory approval from California and Delaware, with a non-binding Microsoft agreement in place but final terms still being negotiated.

🏆 Investors are betting on platform dominance rather than near-term profits, valuing OpenAI's position in AI infrastructure despite projected cash burn of $115 billion through 2029.

OpenAI closed a secondary share sale that pegs the company at $500 billion, nudging it past SpaceX to become the most valuable private firm on the planet. Current and former employees sold roughly $6.6 billion of stock to a buyer group that includes Thrive Capital, SoftBank, Dragoneer, Abu Dhabi’s MGX, and T. Rowe Price. Big number, big message.

What’s equally loud is what didn’t clear. OpenAI authorized more than $10 billion for the tender, so about 36% of the eligible shares stayed put. The company calls that confidence. It could also be caution, fatigue after November’s $1.5 billion SoftBank-led tender, or wariness while the corporate restructuring winds through approvals. All three can be true.

Inside the business, the math is still red. By company figures shared with investors, OpenAI generated about $4.3 billion in the first half of 2025 and burned roughly $2.5 billion doing it. That’s a 58% burn rate. Liquidity, not profit, is the product employees monetized this week.

The premium nobody wants to explain

OpenAI’s last primary round earlier this year valued the company at roughly $300 billion. The new tender prices common shares at a $500 billion headline—about a 67% step-up in half a year. Primary rounds trade governance and protections. Secondaries trade scarcity and access. Still, the gap signals something deeper: late-stage buyers will pay up to get exposure now rather than wait for an exit that may not come soon.

Employees see the spread differently. They’re turning paper into cash at a premium to the last institutional mark, with no firm IPO timeline and a restructuring still in flux. OpenAI’s nonprofit parent has outlined a conversion to a public benefit corporation; a memorandum of understanding with Microsoft is in place, but state approvals are pending. That’s a lot of “ifs.”

From one angle, selling before the corporate plumbing settles is prudent. From another, refusing to sell at $500 billion implies belief in even more upside. Both readings fit. Markets contain multitudes.

The retention play hidden in plain sight

The timing isn’t accidental. Meta has been on a hiring tear, courting senior researchers from OpenAI and elsewhere with packages that reportedly reach nine figures over four years. Apple, Google, and Anthropic are in the same hunt. Talent is the scarce input.

OpenAI’s answer is simple: create liquidity without a resignation letter. A recurring tender turns stock options into bank balances while people stay in their seats. Meta can offer guaranteed cash. OpenAI can offer cash without a job change. Different instruments, same goal.

Plenty of peers use the same tool. SpaceX, Stripe, and Databricks have run periodic secondaries to keep compensation competitive without going public. It’s a pressure valve. It also buys time—time to finish a restructuring, time to negotiate new commercial deals, time to keep shipping.

The restructuring overhang

Under the proposed architecture, the nonprofit would retain control of a new public benefit corporation and receive a large equity stake tied to the operating entity—described by people close to the process as exceeding $100 billion on the tender’s headline valuation. CEO Sam Altman has said he’s “conflicted” about an IPO but sees “tremendous upside” when the company is ready. Both can be true.

The upside isn’t the only variable. Converting governance while preserving mission control invites scrutiny. California and Delaware regulators will want to see how a nonprofit guardian balances fiduciary duty against investor pressure. An ongoing lawsuit from Elon Musk alleging a drift from OpenAI’s founding purpose adds another layer of legal noise. Uncertainty is the tax.

For employees deciding to sell, that uncertainty matters. The 64% uptake says many chose certainty at a rich price. The rest either already took liquidity in November or are betting that the valuation ceiling sits well above today’s print. We’ll know only in hindsight.

The profit horizon, mapped in trillions

OpenAI’s message to investors is consistent: profitability sits late in the decade. Internal scenarios point to cash-flow break-even in 2029, contingent on continued access to massive compute and continued growth in model usage and enterprise contracts. Compute and technical talent are expected to consume the majority of revenue until scale effects kick in. It’s a long march.

Why do buyers still pile in at a $500 billion tag? Because they’re not buying this year’s earnings. They’re buying position. OpenAI has lined up multi-year deals across chips, memory, and cloud capacity and is pushing for a role that looks less like a single app and more like a platform layer. Platform control, if achieved, taxes an ecosystem. Investors are underwriting that possibility.

Rivals are playing the same game—lose money now, win the market later. Anthropic’s annualized revenue has accelerated and, by multiple reports, it is raising at a triple-digit-billion valuation. Google and Meta are scaling foundation models and assistants across their consumer estates. The result is a capital cycle measured in tens to hundreds of billions. Winners will be rare.

What the tender actually signals

Strip out the headlines and the tender says two things. First, OpenAI is comfortable weaponizing private liquidity as a retention tool while it defers public markets. Second, the company can command a price that implies dominant future cash flows even as it burns cash today. That’s not new in Silicon Valley. The scale is.

If you’re an employee, the decision reduces to two questions: Do you trust the restructuring to land cleanly? And do you believe today’s $500 billion will look cheap in three years? There’s no spreadsheet for that. Just appetite for risk.

Why this matters

Private markets as a release valve: Repeat tenders let flagship AI firms stay private for years while still paying people competitively, reshaping how late-stage tech compensates and retains scarce talent.
Price signals over profit: A $500 billion tag on an unprofitable company shows investors are valuing position and platform potential over near-term earnings—raising the stakes for everyone chasing AI scale.

❓ Frequently Asked Questions

Q: What's a secondary share sale and how is it different from normal fundraising?

A: In a secondary sale, employees sell their existing shares to investors—no new money goes to the company. OpenAI gets zero dollars from this $6.6 billion transaction. Primary rounds, by contrast, issue new shares and the cash flows directly to the company. Secondaries create employee liquidity without diluting ownership or changing the cap table structure.

Q: Why would employees choose NOT to sell at a $500 billion valuation?

A: Several reasons. Many likely sold in November's $1.5 billion SoftBank tender and don't need more liquidity. Others may believe the valuation will climb higher before an IPO. Some face tax implications from selling. And uncertainty around the nonprofit-to-PBC restructuring—still awaiting regulatory approval—creates risk that could delay or complicate the transaction.

Q: How does OpenAI's $500 billion valuation compare to public tech companies?

A: At $500 billion, OpenAI would rank among the top 10 most valuable U.S. companies if public. It exceeds Tesla ($650B), matches Berkshire Hathaway ($500B), and dwarfs Netflix ($280B). Only Nvidia ($4T+), Apple, Microsoft, Amazon, and Meta sit materially higher. For a company generating $4.3 billion in half-year revenue, that's roughly 58x run-rate revenue.

Q: How much total capital has OpenAI raised, including this secondary?

A: OpenAI has raised approximately $64 billion in primary funding, including a $40 billion round led by SoftBank in March 2025 (though only $10 billion has been wired, with $20 billion contingent on completing the for-profit conversion). Secondary sales like this $6.6 billion don't add to that total since the money goes to employees, not the company.

Q: How long can OpenAI sustain a $2.5 billion per half cash burn rate?

A: At the end of June 2025, OpenAI held $17.5 billion in cash and securities. Burning $2.5 billion per half means $5 billion annually, giving roughly 3.5 years of runway at current burn—though the company projects burn will increase to $8.5 billion for full-year 2025 and $115 billion cumulatively through 2029, requiring continued capital raises.