OpenAI is on track for a $1 trillion IPO in September 2026. Twelve months ago, the company was valued at less than half that. Now it generates $2 billion in monthly revenue and crossed 1 billion monthly active users, numbers that make the IPO target look like a floor rather than a ceiling.
OpenAI statistics from 2026 capture a company that closed a $110 billion funding round in February at a $730 billion pre-money valuation, then hit $852 billion by March. Annualized revenue reached approximately $25 billion by mid-year, and an advertising pilot alone generated $100 million in recurring revenue within six weeks of launch.
Here is where the money is coming from, how fast the user base is scaling, and what the path to profitability actually looks like.
Key OpenAI Statistics for 2026
OpenAI crossed $852 billion in valuation and $25 billion in annualized revenue by early 2026, with a $122 billion funding round that dwarfed every previous private tech raise on record.
- 900 million weekly active users globally as of February 2026, up from 800 million in October 2025
- $25 billion+ in annualized revenue reached by February 2026, up from $20 billion at the end of 2025, per Sacra’s analysis
- $852 billion post-money valuation reached following the March 2026 funding round, per Sacra and TechMarketBriefs
- $122 billion raised in March 2026, the largest private tech funding round on record
- $190 billion in total funding received as of 2026, with the March 2026 round being the largest single funding round in history, per gradually.ai
- 1 million+ business customers worldwide using OpenAI products, confirmed by OpenAI in November 2025
- 2.5 billion daily queries processed globally across all ChatGPT users
- 36.5% of U.S. businesses adopted OpenAI services by July 2025
- 45.3% market share among daily U.S. mobile app users as of January 2026, down from 69.1% in January 2025, per Apptopia data reported by Fortune
- 330 million daily queries from U.S. users alone, representing roughly 13% of global query volume
- 49% of U.S. companies report using ChatGPT regularly in their workflows, per a ResumeBuilder/Digital Silk survey
- ChatGPT adoption growth rates in the lowest-income countries were more than 4x those in the highest-income countries by May 2025, according to OpenAI’s own research
OpenAI Revenue Statistics and Financial Performance
OpenAI grew from $1.6 billion in revenue in 2023 to $13.07 billion in 2025. It also recorded a $20.92 billion operating loss in the process, spending $1.70 for every dollar it earned.
Period | Revenue | Key Development |
|---|---|---|
Full-year 2023 | $1.6 billion | ARR at $2 billion by year-end |
Full-year 2024 | $3.7 billion | ~$5 billion net loss on same-year revenue |
Full-year 2025 | $13.07 billion | $20.92B operating loss; total costs ~$34 billion |
2026 projection | ~$30 billion | $14B projected loss; cash-flow positive not until 2030 |
OpenAI revenue statistics from mid-2026 show where the money actually comes from, and the composition has shifted faster than the topline suggests:
- 68% from ChatGPT subscriptions (~$17 billion annualized)
- 26% from API consumption (~$6.5 billion annualized)
- 6% from Sora, the advertising pilot, and licensing (~$1.5 billion annualized)
The API share matters. It signals that OpenAI is building a revenue base that does not depend entirely on consumer subscription growth, which tends to plateau. But the cost structure tells a harder story.
Microsoft received $17.2 billion from OpenAI in 2025 for Azure compute alone, per leaked financials verified by the Financial Times. That includes $10.59 billion tied to research and development. Inference costs reached $8.4 billion in 2025, quadrupling year-over-year, and are projected to hit $14.1 billion in 2026.
Adjusted gross margin fell to 33% in 2025 as inference costs surged. By Q1 2026, it recovered to roughly 39%, driven by a 95% drop in inference cost per query since GPT-4 launched in early 2023. Whether that margin recovery holds as model complexity increases is the central question for OpenAI’s path to profitability.
Internal documents project cumulative losses of $44 billion through 2029, with a peak annual loss of approximately $170 billion expected in 2028. Cash-flow positive is not forecast until 2030.
OpenAI Valuation and Funding Round Statistics
The $122 billion OpenAI raised in March 2026 was the largest private fundraise in history. The round’s structure reveals what the headline number does not: conditional commitments, retail access, and a credit backstop pointing toward a company building toward a public listing.
OpenAI’s valuation climbed from $29 billion in 2023 to $852 billion by March 2026, a 29x increase in roughly three years. Total funding across all rounds now stands at approximately $190 billion, and the latest round drew a syndicate wider than any previous private fundraise.
Date | Valuation | Round / Event |
|---|---|---|
2023 | $29 billion | Thrive Capital-led tender offer |
Early 2024 | $80 billion | Secondary share transaction |
October 2024 | $157 billion | Series E funding round |
March 2025 | $300 billion | Series F; $40 billion raised, led by SoftBank ($30 billion) |
October 2025 | $500 billion | Secondary share sale; surpassed SpaceX as most valuable private company |
March 2026 | $852 billion | Series G; $122 billion raised, co-led by SoftBank, a16z, D.E. Shaw, MGX, TPG, T. Rowe Price |
The March 2026 round broke new ground not just in scale but in how the capital was structured:
- $50 billion from Amazon, structured as $15 billion funded at closing plus $35 billion contingent on OpenAI achieving artificial general intelligence or completing an IPO
- $3 billion from individual retail investors through bank channels for the first time, with inclusion in several ARK Invest exchange-traded funds to broaden the shareholder base
- $4.7 billion revolving credit facility, undrawn, supported by a syndicate including JPMorgan Chase, Goldman Sachs, Morgan Stanley, Wells Fargo, and Citi
The conditional structure of Amazon’s commitment is the detail most worth pausing on. Fifty billion dollars sounds unconditional until you read the terms: $35 billion materializes only if OpenAI achieves AGI or goes public. That is not a vote of confidence in today’s technology. It is a bet on a specific outcome, structured to limit downside if that outcome does not arrive.
ChatGPT User Growth Statistics 2026
ChatGPT was nearing 1 billion weekly active users by mid-2026, roughly three years after crossing 100 million. OpenAI reported more than 10% monthly growth as of February 2026, per Reuters, a trajectory that almost never survives at this scale.
Period | Weekly Active Users | Notable Context |
|---|---|---|
November 2023 | 100 million | First major WAU milestone disclosed |
March 2025 | 500 million | 5x growth in 16 months |
August 2025 | ~700 million | YoY growth running at 4x+, per CNBC |
October 2025 | 800 million | 8x from Nov 2023 in under 2 years |
February 2026 | 900 million | Doubled from 400M in Feb 2025 |
Mid-2026 | Nearing 1 billion | Per The Information; OpenAI has not confirmed |
The headline number tracks scale. The data underneath tells a different story about who is driving growth and where it is headed:
- 100 million weekly active users in India alone, per DemandSage
- 52% of identifiable ChatGPT users are now women, up from 37% in January 2024
- 38% of daily queries originate outside North America and Europe, per Zebracat
- 193 million daily active users sending over 2 billion prompts per day, per SecondTalent’s analysis of OpenAI data
- 64% of internet users worldwide have interacted with ChatGPT at least once in the past year, per Zebracat
The fastest adoption gains, per OpenAI’s Q1 2026 Signals data, are in the Dominican Republic, Haiti, and Tanzania. Growth in the lowest-income countries continues to outpace the highest-income ones, a pattern first documented in OpenAI’s September 2025 consumer usage study.
ChatGPT Usage Statistics: How People Actually Use It
Writing made up 36% of all ChatGPT conversations when OpenAI began tracking in June 2024. By June 2025, it had fallen to 24%. Information seeking rose from 14% to 24% over the same period, swapping positions.
The reversal happened inside a broader surge. Daily messages grew from 451 million to 2.6 billion in a single year, per OpenAI’s consumer usage study published as an NBER working paper. The full category breakdown shows where usage is concentrated:
- 29% of conversations involve practical guidance, the single largest category
- 24% seek information, up from 14% a year earlier
- 24% involve writing, down from 36% a year earlier
- 10.2% involve tutoring and teaching, more than double the share for programming
- 4.2% involve computer programming
- 3% involve mathematical calculations
- 1.9% involve relationships and personal reflection
OpenAI’s original work/non-work binary also concealed a third category. When the company added schoolwork as a distinct classification by March 2026, the breakdown revealed how much student use had been folded into the work bucket:
Period | Non-Work | Work-Related | Schoolwork |
|---|---|---|---|
June 2024 | 53% | 47% | Not tracked |
June 2025 | 73% | 27% | Not tracked |
March 2026 | 68.9% | 19.2% | 16.3% |
Education level predicts usage patterns more reliably than demographics. Post-graduate degree holders send 48% work-related messages, compared to 37% for users without a bachelor’s degree, per the NBER analysis. Average sessions run approximately 14 minutes, one of the longest of any major web platform, per Similarweb data.
Web traffic share fell from 76% to 53.9% between June 2025 and May 2026, even as absolute visits held at 5.6 billion monthly, per Similarweb. The mobile app crossed $5.01 billion in total consumer spending by June 2026 across an estimated 1.2 billion downloads, per Appfigures Intelligence, while all-time downloads reached 1.92 billion.
OpenAI Website Traffic Statistics: openai.com vs. chatgpt.com
OpenAI website traffic statistics reveal a more than 10-to-1 gap between the company’s product and its corporate domain. chatgpt.com received 5.51 billion visits in April 2026, ranking it fifth among all websites globally. openai.com recorded roughly 508 million by June, down from 865 million just seven months earlier. Its global rank dropped from #38 to #61 in the same period.
Metric | openai.com | chatgpt.com |
|---|---|---|
Monthly visits | ~508 million | 5.51 billion |
Global website rank | #61 | #5 |
Avg. visit duration | 2 minutes 2 seconds | 6 minutes 5 seconds |
Pages per visit | 2.50 | 3.73 |
Bounce rate | 61.53% | 31.83% |
Primary traffic source | YouTube (42% social) | Direct (76% desktop) |
openai.com’s traffic profile matches a content hub: high bounce rate, short sessions, YouTube driving social referrals. Visitors land on a blog post or news item, read it, and leave. chatgpt.com behaves like a utility. Nearly three-quarters of its desktop traffic arrives through direct navigation, the signature of habitual return visits. One domain publishes. The other delivers.
ChatGPT User Demographics Statistics
44% of U.S. adults have used ChatGPT, more than double the 18% recorded in 2023, per Pew Research Center’s February 2026 survey. The users driving that growth no longer resemble the platform’s early base of young, male, English-speaking tech workers.
Education Level | % of U.S. Adults Who Have Used ChatGPT |
|---|---|
Postgraduate degree | 52% |
Bachelor’s degree | 51% |
Some college | 33% |
High school or less | 18% |
U.S. adults overall (2026) | 44% |
U.S. adults under 30 | 58% |
The education gradient remains the sharpest demographic divide in ChatGPT’s U.S. data. College graduates are nearly three times as likely to have used the platform as adults without a degree, a gap that partly reflects income and occupation. Nearly half of all ChatGPT messages from adults come from users under 26, per OpenAI’s September 2025 NBER consumer usage study. Age appears to compress the education gap at the younger end of the distribution.
- 57% of adults under 50 have used ChatGPT, compared to 28% of those ages 50 and older
- 2.1x more likely to access ChatGPT on desktop for users aged 35+ compared to those aged 18-24, reflecting a sharp device-preference divide
- 53.84% of ChatGPT’s earliest users were male; by mid-2025, that split had reversed, with women now forming the majority, per OpenAI’s NBER study
- 18.06% of ChatGPT traffic originates in the U.S., followed by India at 8.91%, Brazil at 4.93%, Germany at 3.99%, and the U.K. at 3.29%
- 32% of all global ChatGPT users are based in Asia, the largest regional concentration
- 62% of college students report using ChatGPT weekly, with 41% of freelancers in writing, design, and development fields also regular users
- 59% of U.S. teens ages 13-17 have used ChatGPT, per Pew Research Center, with roughly 3 in 10 using AI chatbots on a daily basis
ChatGPT is available in 161+ countries and supports 95 languages. The platform’s reach now extends well beyond the high-income, English-speaking markets where it started. The education divide persists, but it is narrowing as younger and more international cohorts adopt the platform at higher rates.
OpenAI GPT-5 and API Pricing Statistics
GPT-5 undercut every frontier model when it launched at $1.25 per million input tokens in August 2025. Seven months later, OpenAI cut that price in half. Meanwhile, it released two newer models priced at two to four times the original rate. The GPT-5 family now spans a 100x range from cheapest to most expensive input token, and that spread is the strategy.
Model | Input (per M tokens) | Output (per M tokens) | Notes |
|---|---|---|---|
GPT-5 Nano | $0.05 | $0.40 | Summarization and classification |
GPT-5 Mini | $0.25 | $2.00 | 80% less output cost than flagship |
GPT-5 (reduced, May 2026) | $0.625 | Not reported | Halved from launch price; cached input: $0.125 |
GPT-5 (launch, Aug 2025) | $1.25 | $10.00 | Cached input: $0.125/M tokens |
GPT-5.1 (May 2026) | $1.25 | Not reported | Replaced original GPT-5 at same rate |
GPT-5.4 (March 2026) | $2.50 | $15.00 | 1.05M context window; native computer-use |
GPT-5.5 (April 2026) | $5.00 | $30.00 | ~20% higher intelligence per token |
OpenAI spent $5.02 billion on inference with Microsoft Azure in the first half of 2025 alone, per leaked Microsoft financial documents. That already exceeds the $3.767 billion it spent on inference in all of 2024, when total compute costs hit approximately $7 billion. The price compression at the bottom drives adoption. The premium tiers at the top pay for it.
ChatGPT Referral Traffic Statistics
ChatGPT referral traffic converts at 11.4%, outperforming every traditional digital marketing channel, per ECDB research using Similarweb conversion data. Direct traffic sits at 10.2%. Paid search at 9.3%. Organic search, the channel most marketers still prioritize, converts at less than half that rate.
On U.S. transactional sites, ChatGPT-referred visitors convert at 7% versus 5% from Google. They spend 15 minutes on site on average versus 8 from Google and view 12 pages versus 9, per Similarweb’s 2025 Gen AI Landscape report. Adobe Digital Insights found AI-referred visitors converted 31% better than non-AI traffic in January 2026. Those visitors spent 68% more time on site and bounced 33% less.
Traffic Channel | Conversion Rate |
|---|---|
ChatGPT referrals | 11.4% |
Direct traffic | 10.2% |
Paid search | 9.3% |
Organic search | 5.3% |
Email | 4.6% |
Display | 4.0% |
Social media | 3.8% |
A product change in May 2026 reshaped how those referrals flow. ChatGPT replaced footnote citations with clickable brand names in mid-2026. Homepage referrals surged from roughly 26-29% of all referral traffic to about 62-63% by late May, per Similarweb weekly referral data. That level held as a permanent plateau. The shift favors brand familiarity over deep-linked content, meaning sites receiving ChatGPT traffic increasingly see visitors on their homepages rather than specific pages.
ChatGPT referral traffic to third-party websites grew 52% year-over-year from September to November 2025. Over the same period, Google Gemini’s referral traffic grew 388%, per Similarweb data shared with Digiday. Gemini started from a smaller base, but the growth asymmetry signals a competitive shift in how AI platforms distribute traffic to the open web. Roughly 26% of ChatGPT responses now contain an ad. About a third of all ad impressions appear in the first response of a session, per Similarweb’s 2026 Generative AI Landscape report.
GPT-5 Performance and Adoption Statistics
Within one week of GPT-5’s August 2025 launch, reasoning workloads on OpenAI’s platform increased eightfold. By January 2026, only 0.1% of daily ChatGPT users still actively selected GPT-4o. OpenAI retired it and five other legacy models on February 13, 2026.
Benchmark | GPT-5 | GPT-5.2 | GPT-5.5 |
|---|---|---|---|
SWE-bench Verified (coding) | 74.9% | 80.0% | 88.7% |
AIME 2025 (math, no tools) | 94.6% | 100% | — |
GPQA Diamond (PhD-level science) | 88.4% (Pro) | 93.2% (Pro) | 93.6% |
ARC-AGI-1 (abstract reasoning) | — | 90.0% (Pro) | — |
GPT-5.2’s perfect AIME score made it the first model to reach 100% on that benchmark through pure model capability, per OpenAI’s December 2025 documentation. The progression from GPT-4o’s 42.1% to GPT-5’s 94.6% to a perfect score took roughly 18 months. Epoch AI placed the original GPT-5’s capability gains in the same range as the GPT-3 to GPT-4 transition, and subsequent versions widened the gap. GPT-5 with thinking enabled outperforms o3 while using 50 to 80% fewer output tokens across visual reasoning, agentic coding, and graduate-level scientific problem solving.
Where benchmark scores translate into practical outcomes, the gains are measurable:
- 4.8% hallucination rate, down from over 20% in GPT-4o, making GPT-5 approximately 45% less likely to contain a factual error, per OpenAI’s launch documentation
- 70.9% on GDPval, a benchmark measuring well-specified knowledge work across 44 occupations, meaning it beats or ties expert professionals at 11x the speed and less than 1% the cost, per OpenAI’s GPT-5.2 launch data
- 600,000 companies now pay for ChatGPT Enterprise, and more than 92% of Fortune 500 firms use OpenAI products in production, per TechRadar reporting cited by Swarnendu De
GPT-5.4, launched in March 2026, expanded the context window to more than 1 million tokens with native computer-use capabilities, more than doubling the original 400K window. The expansion enables applications that were not feasible at the original scale, from full codebase analysis to multi-document synthesis in a single prompt.
Anthropic overtook OpenAI in U.S. business AI spending for the first time in April 2026: 34.4% versus 32.3%, per Ramp AI Index data. ChatGPT still dominates the consumer market, but that lead is eroding on multiple fronts.
Metric | OpenAI | Anthropic | Google | Grok |
|---|---|---|---|---|
GenAI chatbot market share (Jan 2026) | 67.1% | 8.5% | 14% | |
GenAI chatbot market share (Jul 2026) | 51.3% | 10.3% | 27.7% | |
US mobile app share (2026) | 38.7% | ~17% | 13.5% | |
US business AI spending (Apr 2026) | 32.3% | 34.4% | ||
Enterprise LLM API spend (2025) | 27% | 40% | 21% | |
Enterprise coding (2025) | 21% | 54% |
Six months of movement make the pattern clear:
- ChatGPT’s overall GenAI market share fell from 67.1% to 51.3% between January and July 2026, per FirstPage Sage
- Google Gemini nearly doubled its share from 14% to 27.7% over the same period
- Anthropic wins roughly 70% of head-to-head enterprise deals against OpenAI, per Ramp spending data
- The share of U.S. businesses paying for Anthropic rose from 4% to nearly 25% in a single year
- Claude’s US mobile app share surged from approximately 4% to about 17% between February and June 2026, per Apptopia
OpenAI’s consumer reach remains enormous, but the market is splitting by segment. Anthropic dominates enterprise coding; Google doubled its chatbot share in six months. Grok more than tripled its user base in three months, reaching 117 million monthly active users. Deep platform integration can compress years of user acquisition into months.
OpenAI API Usage Statistics
Codex, OpenAI’s agentic coding agent, reached over 5 million weekly active users by June 2026, up from 2 million three months earlier. By that point, it accounted for more than 90% of the average engineer’s usage on the platform. The shift from conversational AI to agentic tooling is the defining story behind OpenAI’s API usage trajectory.
Agentic models require 5 to 30 times more tokens per task than a standard chatbot query, per Gartner’s March 2026 analysis. That cost structure explains the throughput numbers. OpenAI’s API processed 300 million tokens per minute in 2023. By March 2026, it was 15 billion, a 50x increase driven less by user count than by what each user is asking the platform to do.
Metric | Earlier Baseline | Latest Figure | Change |
|---|---|---|---|
Tokens processed per minute | 300M (2023) | 15B (March 2026) | 50x increase |
Daily API calls | 1.3B (2024) | 2.2B+ (2025) | 69% year-over-year |
Active developers on platform | 2M weekly (2023) | 4M (October 2025) | 2x growth |
Codex weekly active users | 2M (March 2026) | 5M+ (June 2026) | 2.5x in three months |
These throughput numbers track volume. The enterprise data underneath shows what is actually running inside it:
- Reasoning token consumption per enterprise organization increased 320x year-over-year, per OpenAI’s State of Enterprise AI 2025 Report analyzing data across more than 1 million business customers
- The median worker’s output tokens rose at least 10-fold in every job function at OpenAI between November 2025 and June 2026, per an arXiv research paper analyzing internal usage data
- Enterprise makes up more than 40% of OpenAI’s revenue and is on track to reach parity with consumer revenue by the end of 2026, per Chief Revenue Officer Denise Dresser’s April 2026 blog post
- Knowledge workers represent approximately 20% of Codex users and are adopting the platform 3x faster than developers, per Panto AI’s analysis of OpenAI data
Goldman Sachs projects a 24-fold increase in global AI token consumption by 2030, reaching 120 quadrillion tokens per month. OpenAI’s current throughput will need to scale dramatically to keep pace.
OpenAI Enterprise Adoption Statistics 2026
Enterprise ChatGPT adoption has reached a scale that makes the next number hard to square. 74% of companies using AI in the workplace have not yet demonstrated tangible business value from it, per Worklytics data cited by AI Business Weekly.
OpenAI’s enterprise seat count crossed 4 million by Q1 2026 across more than 5,000 customers, with average contracts near $800,000, per Value Add VC. Weekly enterprise messages grew 8 times since November 2024. Deployment is near-universal at the top. Proof of return is not.
Metric | Value | Context |
|---|---|---|
Enterprise seats | 4 million+ | Q1 2026, 5,000+ enterprise customers |
Average contract value | ~$800,000 | Top deals exceed $50M annually (Walmart, JPMorgan) |
Weekly enterprise messages | 8x since Nov 2024 | 30% more per worker year-over-year |
Messages per seat (frontier firms) | 2x median enterprises | Deeper integration into repeatable workflows |
Messages to GPTs (frontier firms) | 7x median enterprises | Custom assistants driving structured usage |
Structured workflow growth | 19x year-to-date | ~20% of messages via tailored assistants |
OpenAI’s State of Enterprise AI 2025 Report also captures the geographic breadth of this growth. Every major non-US market outpaced the 143% global average between November 2024 and November 2025. Australia led at 187% year-over-year, followed by Brazil at 161%, the Netherlands at 153%, and France at 146%. Marketing teams lead use-case adoption at 65% consistent usage for content creation and SEO, per OpenAI’s analysis of 1.5 million enterprise conversations.
More than a quarter of U.S. workers and 45% of those with postgraduate degrees use ChatGPT for work, per OpenAI’s January 2026 report. The distance between deployment speed and value demonstration is where the next phase of enterprise AI gets decided.
OpenAI Data Center and Stargate Statistics
A single gigawatt-scale data center costs $50 billion fully loaded and takes roughly three years to build, per OpenAI CFO Sarah Friar on the All-In Podcast. OpenAI grew from 0.2 GW in 2023 to 1.9 GW by end of 2025, a nearly 10x expansion in two years. The seven Stargate sites now in development would push that to over 9 GW.
Stargate Site | State | Planned Capacity | Key Context |
|---|---|---|---|
Abilene | Texas | 1.2 GW | Flagship site; under construction |
Shackelford County | Texas | 2.0 GW | Largest single site |
Doña Ana County | New Mexico | 2.2 GW | Second-largest site |
Milam County | Texas | 1.2 GW | $1B joint SB Energy investment |
Port Washington | Wisconsin | 1.3 GW | Midwest presence |
Saline Township | Michigan | 1.4 GW | Great Lakes region |
Lordstown | Ohio | <0.3 GW | Smallest site |
The combined 9+ GW across these sites is comparable to New York City’s peak power demand, per Epoch AI’s April 2026 satellite analysis. Altman’s broader vision extends to 30 GW at an estimated $1.4 trillion in total build-out, with a target of deploying 1 GW per week. OpenAI now delivers useful intelligence at costs measured in cents per million tokens, down from under $1 earlier in 2026, per Friar’s January 2026 blog post.
The financial and hardware commitments behind the buildout span three continents and multiple chip architectures:
- $300 billion partnership with Oracle to develop up to 4.5 GW across new Texas, New Mexico, and Midwest sites over five years
- 10 GW of Nvidia systems under a September 2025 strategic partnership, with Nvidia committing to invest up to $100 billion as systems deploy on the Vera Rubin platform, first phase targeted for H2 2026
- 6 GW purchase commitment from AMD for MI300-series accelerators, diversifying the hardware pipeline beyond a single supplier
- $1 billion joint investment with SoftBank in SB Energy to support the 1.2 GW Milam County, Texas campus
- $25 billion Stargate Argentina project in Patagonia, the first Stargate site in Latin America and the largest AI data center in the region
OpenAI also unveiled Jalapeño in June 2026, its first custom inference chip built with Broadcom and designed for LLM workloads. Initial deployment is planned by end of 2026 at gigawatt scale, targeting 10 GW of custom AI accelerators by 2029. The constraint is not demand. It is whether the physical infrastructure can arrive fast enough.
OpenAI Website Traffic and Engagement Statistics
ChatGPT ranked as the fifth most visited website on the planet in May 2026, sitting behind only Google, YouTube, Facebook, and Instagram. It received 5.565 billion visits that month with 24.02% year-over-year growth, per Similarweb. The platform peaked at 6.2 billion visits in October 2025 before settling into a narrower range, but the trajectory from 1.6 billion visits in January 2024 tells the larger story.
Period | Monthly Visits | Context |
|---|---|---|
January 2024 | 1.6 billion | First major monthly visit milestone publicly reported |
July 2024 | 2.4 billion | Surpassed January 2024 figure; per Similarweb |
August 2025 | 5.846 billion | 489 million unique visitors |
October 2025 | 6.2 billion | Highest monthly total reported through early 2026 |
May 2026 | 5.565 billion | Ranked #5 globally; 24.02% YoY growth |
The plateau from October 2025 to May 2026 masks a shift from novelty to routine. Bounce rates dropped from 37.07% in late 2024 to 33.47% by April 2026, per Similarweb data cited by Elfsight, meaning visitors are staying and interacting rather than leaving after a single page.
These visit counts also understate how people actually reach the platform. Approximately 64% of generative AI users now access tools like ChatGPT exclusively through mobile devices, per Sensor Tower data cited by Thunderbit. Website traffic captures only the browser-based fraction of engagement. The real usage footprint is substantially larger than any ranking suggests.
Microsoft OpenAI Partnership Statistics
Microsoft’s $13 billion investment in OpenAI produced a 17.6x paper multiple as of March 2026, with its 26.79% stake valued at approximately $228.3 billion. That return arrived alongside a deal Microsoft chose to fundamentally rewrite.
The April 27, 2026 restructuring removed exclusivity, capped the revenue share, and ended the AGI-linked termination trigger. Microsoft’s IP license converted from exclusive to non-exclusive, and OpenAI gained the freedom to serve any cloud provider. What replaced the original terms is a fixed commercial relationship running through the end of the decade.
Partnership Term | Original Structure | After April 27, 2026 Restructuring |
|---|---|---|
Total Microsoft investment | $13B committed; $11.9B funded as of June 30, 2026 | Unchanged; $3.1B equity-method hit in fiscal Q1 2026 |
Microsoft stake | ~32.5% (prior for-profit structure) | 26.79% diluted; valued at ~$228.3B at March 2026 valuation |
Cloud relationship | Exclusive Azure partnership for API products | Non-exclusive; OpenAI launches first on Azure unless Microsoft cannot support required capabilities |
Revenue share (OpenAI to Microsoft) | ~20% of revenues; ends when AGI verified | Capped at $38B total through 2030; AGI-linked termination removed |
Revenue share (Microsoft to OpenAI) | Microsoft paid OpenAI a share of Azure OpenAI Service and Bing revenues | Microsoft stopped paying OpenAI a revenue share on Azure resales |
IP license | Exclusive through 2032, excluding consumer hardware | Non-exclusive through 2032; other cloud providers may access OpenAI models |
Azure purchase commitment | Not separately disclosed | OpenAI committed to $250B of Azure services (confirmed Oct 2025) |
The revenue share cap reshapes the economics. At the original 20% rate, the payment trajectory implied OpenAI would owe Microsoft roughly $135 billion before 2030. The new cap at $38 billion saves OpenAI an estimated $97 billion compared to that projection. Leaked financial documents reviewed by TechCrunch in November 2025 show OpenAI paid Microsoft $493.8 million in 2024. That figure nearly doubled to $865.8 million in just the first three quarters of 2025. At the original rate, those payments implied revenues of at least $2.5 billion in 2024 and $4.33 billion through nine months of 2025.
- $24.1 billion in revenue Microsoft disclosed from commercial arrangements with OpenAI in fiscal 2026, including Azure purchases and revenue-sharing payments, per Microsoft’s fiscal 2026 Form 10-K
- $6 billion in accounts receivable from OpenAI on Microsoft’s balance sheet as of June 30, 2026
- $228.3 billion in Microsoft’s OpenAI stake represents approximately 8.2% of Microsoft’s $2.78 trillion market capitalization, creating significant concentration in a single private holding
OpenAI Employee and Workforce Statistics
OpenAI generated approximately $5.5 million per employee in 2025, per Epoch AI’s May 2026 analysis. That figure, nearly double Nvidia’s and behind only Anthropic among major AI companies, reframes the $1.37 million median annual compensation the company pays. The headcount behind that revenue grew nearly as fast as the paychecks.
Date / Period | Headcount | Context |
|---|---|---|
November 2023 | ~770 | Pre-scaling baseline |
September 2024 | 3,531 | Nearly 5x increase in under a year |
End of 2024 | 4,467 | OpenAI Foundation year-end figure |
December 2025 | 7,850 | 54.9% year-over-year increase |
Q1 2026 target | ~8,000 | FT reported direct payroll nearer 4,500; gap reflects methodology differences |
The composition and cost structure behind these numbers reveal where OpenAI concentrates its resources:
- $1.5 million average stock-based compensation per worker in 2025, the highest recorded for any major tech startup, per Fortune
- 25th percentile total compensation: approximately $925,000; median: $1,370,000; 90th percentile: $2,792,200
- Research engineers earn $295,000–$530,000 in base salary; software engineers earn $255,000–$590,000
- Engineering makes up roughly 56% of the workforce (~1,480 people), with Business Management (233), Marketing & Product (225), and Human Resources (212) comprising the remainder, per Unify workforce data
- Only 16% of technical roles are held by women, while approximately 44% of all staff identified as women or non-binary in 2025, per OpenAI’s DEI report
The growth trajectory masks a retention problem. OpenAI’s median employee tenure is 1.7 years as of March 2026, down a full year year-over-year, per Revelio Labs. Over 50 senior researchers and a quarter of the company’s leadership have left in two years. Some founded competitors; others joined Meta’s aggressive recruiting pipeline. In August 2025, OpenAI launched a retention-bonus program covering nearly 1,000 employees with payouts from $300,000 to $1.5 million. Active job postings rose 87.5% in 2026 to 945, suggesting the company is replacing losses and expanding at the same time.
OpenAI Profit and Loss Statistics
Leaked audited financial statements verified by the Financial Times show OpenAI’s total costs and expenses in 2024 were $12.48 billion, not the ~$9 billion previously estimated. The company spent $2.37 for every dollar of revenue it generated, producing a loss from operations of $8.78 billion. The gap between the estimate and the audited figure is $3.5 billion, and it reshapes the entire cost picture.
Cost Category | 2024 Amount | Source |
|---|---|---|
R&D spending (training, research, model development) | $7.81 billion | Leaked audited financials |
Inference compute (running models for users) | ~$1.8–2 billion | Epoch AI estimate |
Employee salaries (excl. stock compensation) | >$700 million | Epoch AI / The Information |
Other costs (data, hosting, sales, marketing) | ~$2 billion | Derived from total |
Total costs & expenses | $12.48 billion | Leaked audited financials, Fortune |
Loss from operations | $8.78 billion | Leaked audited financials |
R&D spending alone at $7.81 billion exceeded the company’s entire 2024 revenue. That is the structural reality underneath the topline growth story. But the more consequential numbers sit in the forward trajectory, where OpenAI and Anthropic start from nearly identical positions and end up in different financial universes. Both companies burned through roughly 70% of their revenue in 2025, OpenAI on a projected $9 billion spend against $13 billion in sales, Anthropic on approximately $3 billion against $4.2 billion. From that similar starting point, the paths diverge dramatically. Anthropic projects an operating profit of approximately $559 million for Q2 2026, per leaked financial documents analyzed by MLQ. OpenAI expects to consume roughly 14 times more cash than Anthropic before reaching profitability, per financial documents reviewed by the Wall Street Journal. OpenAI projects operating losses of approximately $74 billion in 2028, representing roughly 75% of that year’s revenue, driven primarily by escalating computing costs. By 2030, the company targets approximately $200 billion in annual revenue, a figure that would need to absorb the $74 billion annual loss still projected two years earlier. One company is already projecting a quarterly profit. The other is projecting an annual loss in 2028 that is roughly six times its 2024 cost base.
OpenAI Future Revenue Projections and Growth Forecasts
OpenAI increased its five-year revenue forecast by 27% in February 2026, now targeting approximately $280 billion in total revenue through 2030. FutureSearch’s July 2026 expert forecast puts mid-2027 ARR at a median of $44 billion. That is roughly a third below OpenAI’s internal target of $62 billion for the same period. The distance between those two numbers is the most honest measure of what these projections actually contain.
Year / Period | OpenAI Internal Target | Independent Forecast | Notes |
|---|---|---|---|
2026 (annual) | $30 billion | — | Bloomberg, Feb 2026 |
Mid-2027 (ARR) | $62 billion | $44 billion median (range: $30B–$63B) | The Information; FutureSearch, July 2026 |
2029 (annual) | ~$125 billion | — | CNBC, Sep 2025; prior target contributing to 2030 cumulative |
2030 (cumulative) | ~$280 billion total | Break-even not projected until 2031+ | Bloomberg; BNP Paribas, HSBC, FutureSearch |
Beneath these revenue targets, the cash burn has also been revised sharply upward. OpenAI now projects cumulative spending of $665 billion through 2030, with annual burn reaching $57 billion in 2027, per The Information and CNBC. HSBC estimates the company will need at least $207 billion in additional funding to sustain that trajectory. Training costs alone account for $440 billion of the projected spend. The advertising business, projected at $100 billion by 2030 after generating $100 million in ARR within six weeks of launch, is one response to the financing challenge. Whether a new ad channel can scale fast enough against that spend is the question no forecast answers.
Sources
- https://aibusinessweekly.net/p/openai-statistics (2026-06-25)
- https://tech-insider.org/openai-110-billion-funding-round-2026 (2026-04-02)
- https://valueaddvc.com/blog/openai-revenue-2026-20b-arr-4b-month-path-to-profitability (2026-06-25)
- https://www.zacks.com/featured-articles/781/openai-ipo (2026-06-08)
- https://tech-insider.org/openai-122-billion-funding-round-852-billion-valuation-2026 (2026)
- https://www.gradually.ai/en/openai-statistics (2026)
- https://aiafterhours.substack.com/p/openai-vs-anthropic-the-121-billion (2026-06-20)
- https://blog.herlein.com/post/ai-inference-costs-reality-check (2026-06-10)
- https://mlq.ai/news/leaked-openai-financials-reveal-21b-operating-loss-on-13b-revenue-in-2025 (2026-06-20)
- https://finance.yahoo.com/news/openais-own-forecast-predicts-14-150445813.html (2026-06-20)
- https://x.com/saastr/status/2026045904189297014 (2026-06-15)
- https://openai.com/index/accelerating-the-next-phase-ai (2026-03-31)
- https://pinggy.io/blog/openai_anthropic_funding_history (2026-06-01)
- https://techcrunch.com/2026/03/31/openai-not-yet-public-raises-3b-from-retail-investors-in-monster-122b-fund-raise (2026-03-31)
- https://www.getpanto.ai/blog/chatgpt-statistics (2026-07-01)
- https://www.secondtalent.com/resources/chatgpt-statistics (2026-06-01)
- https://thedigitalelevator.com/blog/chatgpt-statistics (2026-02-01)
- https://www.demandsage.com/chatgpt-statistics (2026-08-01)
- https://www.zebracat.ai/post/chatgpt-usage-statistics (2026-06-01)
- https://openai.com/signals/research/2026q1-update (2026-05-11)
- https://momenticmarketing.com/blog/top-ai-chatbots (2026-05-26)
- https://appfigures.com/resources/insights/chatgpt-fastest-app-5b-mobile (2026-06-02)
- https://www.blog.udonis.co/mobile-marketing/mobile-apps/chatgpt (2026-06-01)
- https://www.searchenginejournal.com/chatgpt-study-1-in-4-conversations-now-seek-information/556104 (2025-09-01)
- https://www.nber.org/system/files/working_papers/w34255/w34255.pdf (2025-09-01)
- https://mangothrive.com/how-many-chatgpt-users-are-there (2026-06-01)
- https://www.omnibound.ai/blog/chatgpt-user-statistics (2026-06-01)
- https://www.tdmp.co.uk/insights/who-using-chatgpt-2025-demographic-exploration (2025-10-01)
- https://thunderbit.com/blog/chatgpt-statistics-user-growth (2026-01-01)
- https://explodingtopics.com/blog/chatgpt-users (2025-12-01)
- https://ppc.land/chatgpt-drops-to-52-7-as-claude-triples-its-ai-traffic-share (2026-06-11)
- https://analytics.explodingtopics.com/website/openai.com (2026-06-01)
- https://www.pewresearch.org/internet/2026/06/17/americans-and-ai-2026-chatbots-smart-devices-and-views-on-impact (2026-06-17)
- https://youthrex.com/wp-content/uploads/2026/03/PI_2025.12.09_Teens-Social-Media-AI_REPORT.pdf (2025-12-09)
- https://thunderbit.com/blog/chatgpt-adoption-statistics-business (2026-01-01)
- https://openai.com/business/guides-and-resources/the-state-of-enterprise-ai-2025-report (2025-12-17)
- https://thenetworkinstallers.com/blog/ai-in-the-workplace-statistics (2026-01-01)
- https://boothassociatesllc.com/ai-statistics-small-business-2026.html (2026-04-01)
- https://www.wheresyoured.at/oai_docs (2025-09-01)
- https://epoch.ai/data-insights/openai-compute-spend (2025-10-10)
- https://muneebdev.com/openai-api-pricing-2026 (2026-03-17)
- https://apidog.com/blog/gpt-5-5-pricing (2026-06-17)
- https://intuitionlabs.ai/articles/llm-api-pricing-comparison-2025 (2026-01-01)
- https://blockspace.media/insight/openai-leaked-financials-ai-scaling-costs-2025 (2025-12-01)
- https://www.similarweb.com/blog/marketing/geo/zero-click-marketing (2026-06-10)
- https://www.similarweb.com/blog/marketing/geo/gen-ai-stats (2026-07-29)
- https://www.panxo.com/blog/agentic-commerce-chatgpt-conversion-rates-outperform-every-channel (2026-01-01)
- https://www.digitalapplied.com/blog/ai-referral-traffic-share-2026-gemini-chatgpt-geo-analysis (2026-06-15)
- https://onyx.app/insights/best-llms-2026 (2026-07-20)
- https://www.digitalapplied.com/blog/gpt-5-2-codex-openai-model-guide-2026 (2026-01-31)
- https://docs.apiyi.com/en/news/gpt-5-2-launch (2025-12-11)
- https://www.gend.co/blog/gpt-5-for-work (2026-01-01)
- https://www.swarnendu.de/blog/gpt-5-for-enterprise-the-step-by-step-implementation-guide (2026-01-01)
- https://fatjoe.com/blog/grok-ai-stats (2026-07-01)
- https://ramp.com/leading-indicators/ai-index-may-2026 (2026-05-01)
- https://firstpagesage.com/reports/top-generative-ai-chatbots (2026-08-03)
- https://ramp.com/leading-indicators/ai-index-march-2026 (2026-03-01)
- https://openai.com/index/next-phase-of-enterprise-ai (2026-04-08)
- https://arxiv.org/html/2606.26959v1 (2026-06-24)
- https://www.getpanto.ai/blog/openai-statistics (2026-07-01)
- https://www.cockroachlabs.com/blog/agentic-ai-costs-at-scale (2026-03-01)
- https://www.index.dev/blog/llm-enterprise-adoption-statistics (2026-01-01)
- https://aibusinessweekly.net/p/chatgpt-statistics (2026-06-25)
- https://openai.com/business/guides-and-resources/chatgpt-usage-and-adoption-patterns-at-work (2026-01-22)
- https://timesofindia.indiatimes.com/technology/tech-news/openai-cfo-sarah-friar-declares-priority-in-2026-is-closing-the-gap-between-what-ai-now-makes-possible-and-how-/articleshow/126805158.cms (2026-01-19)
- https://epoch.ai/publications/openai-stargate-where-the-us-sites-stand (2026-04-17)
- https://investor.nvidia.com/news/press-release-details/2025/OpenAI-and-NVIDIA-Announce-Strategic-Partnership-to-Deploy-10-Gigawatts-of-NVIDIA-Systems/default.aspx (2025-09-22)
- https://finance.biggo.com/news/b924f504c62ac6ee (2026-06-01)
- https://www.axios.com/2025/10/28/openai-1-trillion-altman (2025-10-28)
- https://openai.com/index/openai-broadcom-jalapeno-inference-chip (2026-06-24)
- https://intuitionlabs.ai/articles/oracle-openai-300b-deal-analysis (2026-04-11)
- https://en.wikipedia.org/wiki/Stargate_LLC (2025-10-10)
- https://www.similarweb.com/blog/research/market-research/most-visited-websites (2026-06-10)
- https://elfsight.com/blog/chatgpt-usage-statistics (2026-04-01)
- https://www.incremys.com/en/resources/blog/chatgpt-statistics (2026-01-01)
- https://www.techi.com/microsoft-openai-13b-investment (2026-04-04)
- https://www.neowin.net/news/microsoft-reveals-it-generated-241-billion-in-revenue-from-openai-in-fiscal-2026 (2026-07-17)
- https://aimagazine.com/news/openai-caps-microsoft-revenue-share-at-us-38bn-in-new-deal (2026-05-13)
- https://sqmagazine.co.uk/how-many-people-work-at-openai (2026-01-01)
- https://www.leoniscap.com/research/openai-building-the-everything-platform-in-ai (2025-08-01)
- https://www.makerstations.io/openai-employee-statistics (2026-01-01)
- https://www.reveliolabs.com/companies/openai-foundation/employees (2026-03-01)
- https://fortune.com/2026/06/16/openai-financials-leaked-losses-revenue-profit (2026-06-16)
- https://www.investing.com/news/company-news/anthropic-set-to-become-profitable-before-openai-amid-ai-race-4349418 (2025-11-11)
- https://fortune.com/2025/11/12/openai-cash-burn-rate-annual-losses-2028-profitable-2030-financial-documents (2025-11-12)
- https://mlq.ai/news/openai-revises-projections-upward-with-112-billion-extra-cash-burn-by-2030 (2026-02-21)
- https://www.bloomberg.com/news/articles/2026-02-20/openai-forecasts-its-revenue-will-top-280-billion-in-2030 (2026-02-20)
- https://www.reuters.com/commentary/breakingviews/openais-ipo-has-sam-altman-problem-2026-05-21 (2026-05-21)
- https://www.axios.com/2026/04/09/openai-100-billion-in-ad-revenue (2026-04-09)
- https://www.cnbc.com/2026/02/20/openai-sees-compute-spend-of-around-600-billion-by-2030.html (2026-02-20)
- https://futuresearch.ai/openai-financial-forecast (2026-07-13)
- https://fortune.com/2025/11/26/is-openai-profitable-forecast-data-center-200-billion-shortfall-hsbc (2025-11-26)
- https://keyholesoftware.com/enterprise-agentic-ai-market-2026 (2026-06-01)
