OpenAI Just Raised $122 Billion — The Largest Private Funding Round in History

TL;DR

OpenAI closed a $122 billion funding round at a post-money valuation of $852 billion. Amazon put in $50 billion, Nvidia $30 billion, SoftBank $30 billion. It’s the largest private capital raise in corporate history. The money will fund a “superapp” merging ChatGPT, Codex, and the Atlas browser, plus an IPO expected in late 2026 or early 2027. OpenAI is generating $2 billion per month in revenue but still isn’t profitable and expects to lose $14 billion this year.

The Numbers Behind the Raise

The round initially targeted $110 billion. It ended at $122 billion after OpenAI opened participation to retail investors through bank channels for the first time, pulling in $3 billion from individuals. Private companies don’t typically let small investors in before an IPO, and the reason here is obvious: build a shareholder base before the listing.

Here’s how the major investments break down:

The $852 billion valuation makes OpenAI worth more than all but a handful of publicly traded companies. That’s roughly Berkshire Hathaway territory, except Berkshire actually makes money. OpenAI lost money last year and expects to lose $14 billion in 2026.

OpenAI also expanded its revolving credit facility to $4.7 billion, backed by JPMorgan, Citi, Goldman Sachs, Morgan Stanley, and Wells Fargo. The facility is currently undrawn. It’s a safety net, not active spending.

Where the Money Goes

Sam Altman and his team have laid out three main priorities for the cash.

First, the superapp. Fidji Simo, OpenAI’s CEO of Applications (hired from Instacart in May 2025), wrote an internal memo in March calling the multi-product strategy a distraction. ChatGPT, Codex, and the Atlas browser are being merged into a single desktop application. The AI can jump between tools autonomously: research something with Atlas, write code with Codex, explain it through ChatGPT, all without the user switching apps. Mobile ChatGPT stays separate.

Second, compute and infrastructure. OpenAI burns through GPU clusters at a rate that makes the $14 billion projected loss make sense. Training frontier models and serving 900 million weekly active users requires data center capcity that doesn’t exist yet. A chunk of this money will go directly to building and leasing that capacity.

Third, headcount. The company plans to nearly double its workforce to 8,000 employees by end of 2026, hiring across engineering, research, sales, and a new “technical ambassador” program.

The Product Strategy Pivot

This raise comes at a strange moment for OpenAI. Over the past two weeks, the company has killed or shelved three products:

• Sora, the video generation tool, shut down entirely. Disney’s reported $1 billion deal collapsed with it.
• Instant Checkout, an in-chat shopping feature, quietly pulled.
• Adult Mode, shelved after safety advisors flagged severe risks.

Simo’s memo was blunt about why: “spreading team energy across too many standalone apps was slowing momentum and hurting quality control.” She told employees they couldn’t afford “side quests” given how fast Anthropic was gaining ground.

And Anthropic is gaining ground. Claude overtook ChatGPT as the most downloaded app in the US in March 2026. Claude Code is generating roughly $2.5 billion in annualized revenue — more than double Codex’s approximately $1 billion. OpenAI’s coding tool, which was supposed to be its enterprise growth engine, is losing the race to a smaller competitor.

The superapp strategy is the response: stop building scattered products, consolidate into one thing that does everything. It’s WeChat logic applied to AI.

The IPO Shadow

Everyone involved knows this round is pre-IPO positioning. OpenAI is targeting a listing in Q4 2026 or Q1 2027, likely on NASDAQ.

Several moves point to it:

• Letting retail investors into the round builds a shareholder base before the listing
• ARK Invest is adding OpenAI to several ETFs, giving more people exposure to the stock pre-IPO
• The company has restructured from its original nonprofit model to a traditional corporate structure with clear equity rights
• S&P Global, FTSE Russell, and NASDAQ are all considering “fast-track” index inclusion rules that would add OpenAI to major indices within days of listing, skipping the usual 12-month waiting period

The board is reportedly anxious about timing. Anthropic is also exploring an IPO, potentially as early as Q4 2026, with a target valuation around $60 billion, potentially anchored by the unreleased Claude Mythos model. OpenAI’s worry: if Anthropic lists first, it could absorb the pent-up retail demand for AI exposure. There’s a window, and both companies are racing toward it.

The Bull and Bear Case

The case for OpenAI is scale. 900 million weekly active users, $2 billion in monthly revenue, 50 million subscribers, enterprise revenue growing toward 40% of total. It’s the default AI brand for most consumers. If the superapp works and agentic capabilities convert free users to paid, the revenue trajectory could justify the valuation.

The case against is the math. OpenAI expects to lose $14 billion this year. It won’t be profitable until 2030 by its own projections. And the investor list has a weird circularity to it: Amazon ($50 billion in) is also a compute supplier and competitor with its own models. Nvidia ($30 billion in) sells the hardware that makes up OpenAI’s largest cost. The companies profiting from OpenAI’s spending are financing that spending.

Then there’s the competitive picture. Anthropic is winning the developer market. Google’s Gemini 3.1 Pro leads on most benchmarks. Meta’s open-source models keep eroding the willingness to pay for proprietary ones. xAI’s Grok 4.20 is taking a multi-agent approach that could differentiate. The moat, if it exists, is brand recognition and user habit, not technical superiority.

What This Means for the AI Market

A few things to watch going forward.

If OpenAI goes public at $852 billion or above, it sets the reference point for every AI company valuation. Anthropic’s rumored $60 billion IPO suddenly looks modest. Every Series B AI startup’s pitch deck just got more aggressive.

There’s also the profitability question. OpenAI is openly saying it won’t be profitable for four more years while raising $122 billion. The last time the tech industry saw this pattern (massive capital raises, no profitability timeline, growth-at-all-costs) it was 2020-2021, and the correction was brutal. The question isn’t whether AI is real, but whether current spending levels can generate returns before investor patience runs out.

And for developers specifically: the superapp strategy means OpenAI is trying to own the entire workflow, not just code completion. If it works, standalone AI browsers, separate coding agents, and individual AI assistants become features inside a single OpenAI product. That’s bad news for every startup building one piece of the AI toolchain.

FAQ

How much is OpenAI worth now?

$852 billion post-money, based on the just-closed funding round. That’s up from $300 billion in October 2024 and $157 billion in early 2024.

When will OpenAI go public?

The company is targeting Q4 2026 or Q1 2027 for a NASDAQ listing. No S-1 has been filed yet.

Is OpenAI profitable?

No. The company generates $2 billion per month in revenue but projects a $14 billion loss in 2026. It doesn’t expect profitability until 2030.

Who invested in OpenAI’s latest round?

Amazon ($50B), Nvidia ($30B), SoftBank ($30B), Andreessen Horowitz, Microsoft, D.E. Shaw Ventures, and for the first time, $3 billion from retail investors.

What is OpenAI’s superapp?

A planned desktop application merging ChatGPT, Codex (coding), and the Atlas browser into a single product with agentic capabilities. Mobile ChatGPT will remain separate.

Bottom Line

$122 billion is a staggering number for a company that’s never turned a profit. The bet is on owning the default AI interface the same way Google owns search. But Google was profitable when it made its platform bet. OpenAI is making the same play while burning $14 billion a year and racing a competitor that’s eating its developer market from underneath.

The raise buys time. Probably two to three years of runway at current burn rates. The IPO will buy more. But at some point the revenue has to catch the spending, and nobody has explained exactly how that happens by 2030. They’re betting it will. That’s a $122 billion bet.