OpenAI Revenue, Valuation, and Funding in 2026

OpenAI says it reached $25 billion in annualized revenue in February 2026, up from $20 billion at the end of 2025. That same year, the company landed a $122 billion investment round at an $852 billion valuation, which is the kind of number that forces everyone else in AI to recalculate their own plans.

The pace is even more striking when you line it up with usage: weekly active users climbed to 910 million, and paying business users passed 9 million. This is no longer a research lab with a popular chatbot. It is a giant software business with the cost structure, capital needs, and investor expectations that come with scale.

Revenue is growing fast, but the bill is huge

OpenAI’s revenue story is simple on the surface and messy underneath. The company is pulling in money from ChatGPT subscriptions, API usage, and enterprise deals, while still spending heavily on inference to keep the models running. Sacra’s estimate puts annualized revenue at $25 billion in February 2026, and CFO Sarah Friar confirmed $20 billion for 2025, after $6 billion in 2024 and $2 billion in 2023.

That growth is real, but so is the burn. OpenAI projected $8.4 billion in inference costs in 2025 and $14.1 billion in 2026. It also projected cash burn of about $9 billion in 2025 and $17 billion in 2026, with cash-flow positivity not expected until 2030.

• Annualized revenue: $25B in Feb. 2026
• 2025 revenue: $20B, confirmed by Sarah Friar
• 2024 revenue: $6B
• 2023 revenue: $2B
• Weekly active users: 910M
• Paying business users: 9M+

The mix matters as much as the total. OpenAI says subscriptions make up the bulk of revenue, with API usage and enterprise sales filling in the rest. That is a strong business, but it is also one where every new feature can make the compute bill climb faster than the top line.

OpenAI has also started pushing deeper into workflows that make users stick around longer. Spreadsheet editing, presentation editing, and tools like ChatGPT for Excel are designed to move the product from “I ask it questions” to “I do my work here.” That is where subscription revenue gets stickier.

The valuation jump changed the conversation

The funding history explains why investors are treating OpenAI like a category unto itself. In March 2025, the company raised $40 billion at a $300 billion post-money valuation. By October 2025, current and former staff sold about $6.6 billion of stock in a secondary deal at a $500 billion valuation. Then in April 2026, OpenAI announced another $122 billion in new investment at an $852 billion valuation, led by SoftBank with Andreessen Horowitz, D.E. Shaw Ventures, MPX, and TPG.

That sequence is unusual even by private-market standards. The company moved from $300 billion to $500 billion to $852 billion in about a year, while still operating with heavy cash burn and a business model that depends on enormous infrastructure spend. Investors are clearly betting that OpenAI can become the default interface for AI work, consumer chat, and developer tooling all at once.

“We think OpenAI is the most important company in the world right now.” — Sam Altman, speaking at the Y Combinator AI Startup School in 2023.

The quote matters because it captures the pitch OpenAI keeps selling to investors: this is not a single product company. It is a platform for consumer AI, developer APIs, and agentic software, with the chance to sit in the middle of a huge amount of digital work. Whether that justifies the valuation is a different question, but that is the bet.

OpenAI’s recapitalization in October 2025 also changed the ownership picture. The for-profit subsidiary became OpenAI Group PBC, while the nonprofit was renamed the OpenAI Foundation and kept control. The Foundation’s stake was valued around $130 billion, and Microsoft’s stake in the PBC was valued around $135 billion, or roughly 27% on an as-converted diluted basis.

Microsoft, revenue share, and the IPO setup

OpenAI’s relationship with Microsoft is still one of the most important parts of the story. Under a renegotiated partnership in October 2025, OpenAI agreed to pay Microsoft 20% of total revenue through 2032, with deferred payments weighted toward later years. OpenAI projected more than $13 billion in total revenue share across 2026 and 2027, mostly to Microsoft.

That is a big slice of the pie, but it also buys OpenAI time and access to the infrastructure it needs. The company can grow now and settle up later, which is useful when your compute bill is measured in billions and your product roadmap keeps expanding.

• Microsoft revenue share: 20% through 2032
• Projected revenue share to Microsoft and others: more than $13B in 2026–2027
• Cash burn projected: $9B in 2025, $17B in 2026
• Cash-flow positive target: 2030
• Potential IPO timing: H2 2026 filing, 2027 listing

OpenAI is also preparing for a public listing. It hired Cynthia Gaylor, the former CFO of DocuSign, as its first head of investor relations, which is a very public sign that the company wants a cleaner story for Wall Street. Internal targets reportedly include a filing in the second half of 2026 and a 2027 IPO, with a possible valuation near $1 trillion.

That would put OpenAI in a tiny club, but the public markets will ask harder questions than private investors have so far. They will want to know whether revenue can keep outrunning compute costs, whether enterprise adoption can deepen without margin damage, and whether ChatGPT can keep growing once the novelty fades.

Products are becoming the revenue engine

OpenAI’s product lineup is broader than most people remember. The company has ChatGPT, DALL·E, Whisper, the OpenAI API, Codex, and Sora. Each one points at a different monetization path, but ChatGPT still does most of the heavy lifting.

ChatGPT has become a real-time, multimodal assistant used by more than 900 million people weekly. OpenAI says 73% of chats are non-work related, and nearly half of conversations come from users aged 18 to 25. Women make up the majority of the user base. That is a consumer product with unusually broad reach, and it gives OpenAI a distribution base that most enterprise AI vendors would kill for.

Codex is where the company is trying to win developer mindshare. It is a coding environment built around the terminal and a native Mac app, aimed at interactive software work like refactoring, debugging, and test generation. OpenAI is also pushing into security with Codex Security, which scanned more than 1.2 million commits during a 30-day private beta and found 792 critical issues plus 10,561 high-severity findings.

The API side matters too. OpenAI says developers can build persistent agents with function calling and the Assistants API, while model access keeps getting faster and cheaper. That matters because API usage is one of the cleanest ways for OpenAI to turn model capability into recurring revenue outside the consumer app.

Here is the practical comparison:

• ChatGPT: consumer and enterprise subscriptions, biggest revenue driver
• OpenAI API: usage-based developer revenue, more technical but stickier in embedded products
• Codex: coding workflow monetization, aimed at engineers
• Sora: early-stage media and entertainment monetization, still experimental

The product strategy is clear. OpenAI wants to be the place where people chat, code, analyze, and generate media. If that works, the company becomes far more than a chatbot vendor. If it doesn’t, the valuation math gets uncomfortable fast.

What to watch next

The next year should answer one question: can OpenAI keep scaling revenue faster than it scales compute costs? If the company keeps adding paying users, grows enterprise contracts, and keeps the IPO timetable intact, the market will treat the $852 billion valuation as a stepping stone rather than a peak.

My read is simple. Watch the margin trend, the 2026 filing window, and how much of OpenAI’s growth comes from subscriptions versus enterprise deals. If the company can push cash burn down while revenue keeps climbing, the public listing could be one of the biggest software debuts in years. If not, the market will force a reset long before 2030.

For more context on how AI companies are pricing their future, see our coverage of Anthropic’s funding surge and OpenAI Codex Security.