ChatGPT Statistics (2026): Users, Revenue, Downloads, and Usage
You need a ChatGPT statistic for a deck, report, or article, but the first number you find may be a user count, a website-visit estimate, an app-install total, or a revenue run rate.
Those numbers describe different things. Treating them as interchangeable makes the statistic less useful.
The short answer: ChatGPT reached approximately 1 billion weekly active users in July 2026, according to an OpenAI spokesperson quoted by The Verge. OpenAI had previously reported more than 900 million weekly active users and more than 50 million consumer subscribers in February 2026. Separately, CNBC reported in August 2026 that OpenAI’s company-wide annualized revenue pace had exceeded $40 billion.
This guide separates ChatGPT statistics by date, scope, and source so you can use them without blurring the metric.
ChatGPT at a Glance in 2026
Statistic | Figure | Date and scope | Source |
Weekly active users | Approximately 1 billion | July 2026 | OpenAI spokesperson reported by The Verge |
Consumer subscribers | More than 50 million | February 27, 2026 | OpenAI |
OpenAI annualized revenue pace | More than $40 billion | August 2026 | CNBC reporting |
ChatGPT consumer usage that is work-related | Approximately 30% | 2025 consumer-plan study | OpenAI/NBER research |
ChatGPT consumer usage that is non-work | Approximately 70% | 2025 consumer-plan study | OpenAI/NBER research |
Peak monthly ChatGPT.com visits in 2025 | Approximately 6.17 billion | October 2025 | Similarweb estimate |
Cumulative mobile consumer spending | $2 billion | May 2023 through August 2025 | Appfigures estimate reported by TechCrunch |
Mobile app downloads | Approximately 690 million | Through August 2025 | Appfigures estimate reported by TechCrunch |
Mobile spending per download | $2.91 | Through August 2025 | Appfigures estimate reported by TechCrunch |
The table is a reference list, not a single growth calculation.
A weekly active user is a person active during a week. A web visit is an observed website session. An app download is an installation. Consumer spending measures purchases through mobile app stores. An annualized revenue pace projects current revenue over 12 months.
Those metrics should not be combined as if they measured the same population.
Get the Mobile Testing Playbook Used by 800+ QA Teams
Discover 50+ battle-tested strategies to catch critical bugs before production and ship 5-star apps faster.
How to Read ChatGPT Statistics
Before using a ChatGPT statistic, check four things: the metric, population, date, and source.
That distinction is the difference between a useful data point and a misleading headline.
What does the metric measure?
Weekly active users (WAUs): users active during a seven-day period under the reporting methodology.
Monthly active users (MAUs): users active during a month. MAUs and WAUs should not be placed on one trend line as though they are equivalent.
Website visits: estimated visits to a web property. One person can generate many visits.
Downloads: app installations. A download does not establish continued activity or retention.
Consumer subscribers: paying consumer customers at a reported point in time.
Consumer spending: money spent through mobile app stores. It does not capture every way OpenAI makes money.
Annualized revenue run rate: the current company revenue pace extended over a year. It is not revenue already recognized for a completed financial year.
Who supplied the number?
There are three useful source tiers for this topic:
First-party figures: numbers published directly by OpenAI.
Company figures reported by established publications: for example, a statistic supplied by an OpenAI spokesperson to The Verge or company financial information reported by CNBC.
Independent measurement: estimates from analytics providers such as Similarweb and Appfigures.
A fourth category, secondary statistics roundups, can help you find a source but should not replace the underlying evidence when that evidence is available.
The same discipline applies when comparing adoption claims across AI products. Quash's AI testing statistics and adoption data separates experimentation, individual tool usage, production deployment, and enterprise-wide adoption because those measurements answer different questions.
ChatGPT Users: From 500 Million to 1 Billion
ChatGPT's reported weekly active user count grew rapidly through 2025 and into 2026.
Date | Reported weekly active users | Attribution |
End of March 2025 | 500 million | OpenAI figure reported by TechCrunch |
August 2025 | Approaching 700 million | OpenAI figure reported by TechCrunch |
October 6, 2025 | More than 800 million | OpenAI at DevDay 2025 |
February 27, 2026 | More than 900 million | OpenAI |
July 2026 | Approximately 1 billion | OpenAI spokesperson reported by The Verge |
OpenAI CEO Sam Altman announced at DevDay on October 6, 2025 that more than 800 million people were using ChatGPT each week. OpenAI's DevDay 2025 page also reported 4 million developers building with OpenAI and API throughput of 6 billion tokens per minute at that point.
On February 27, 2026, OpenAI said ChatGPT had grown to more than 900 million weekly active users. The company simultaneously reported more than 50 million consumer subscribers. (OpenAI)
The next major milestone came in July.
An OpenAI spokesperson told The Verge that ChatGPT reached approximately 1 billion weekly active users in July 2026.
OpenAI also said in August that more than 1 billion people were putting ChatGPT to work, while its Signals dataset covers individual Free, Go, Plus, and Pro accounts. (OpenAI)
The 1 billion figure needs context.
Moving from 900 million weekly users in February to approximately 1 billion in July took about five months. That is still enormous absolute growth, but it was slower than several earlier reported milestones.
Dates therefore matter as much as endpoints when interpreting the growth curve.
ChatGPT Subscribers and OpenAI Revenue
OpenAI reported more than 50 million consumer subscribers on February 27, 2026. (OpenAI)
That figure should not be divided by the later 1 billion weekly-active-user count to manufacture a subscriber conversion rate.
The two statistics come from different dates, and the public reporting does not provide enough methodology to establish that the populations can be compared directly.
Revenue requires an additional distinction.
CNBC reported in August 2026 that OpenAI's annualized revenue pace had exceeded $40 billion. (CNBC)
That is OpenAI company revenue, not a standalone ChatGPT revenue figure.
OpenAI earns money through consumer subscriptions, business products, API usage, and other commercial activity. The $40 billion figure should therefore not be described as "$40 billion in ChatGPT revenue."
CNBC's reporting also described enterprise revenue overtaking consumer revenue as OpenAI's business mix shifted further toward organizations.
A revenue run rate is also not the same as completed annual revenue. It describes the pace at which the company was generating revenue at that point in time.
For a broader look at why AI-product adoption and company revenue do not automatically establish product quality, see Quash's analysis of vibe coding adoption, tools, and verification risks.
Scale tells you that people are using a product. It does not independently establish whether every generated output is correct, safe, or ready to ship.
How People Use ChatGPT
Unlike many widely circulated ChatGPT usage claims, several behavioural statistics can be traced directly to research published by OpenAI and the National Bureau of Economic Research.
OpenAI's 2025 study analyzed approximately 1.5 million consumer ChatGPT conversations using privacy-preserving automated classification. Researchers did not manually read user messages. (OpenAI)
The study found that approximately:
70% of consumer ChatGPT usage was non-work-related
30% was work-related
75% of conversations focused on practical guidance, seeking information, and writing
49% of messages were classified as "Asking"
40% were classified as "Doing"
11% were classified as "Expressing"
These percentages apply to the study's consumer-plan population and methodology. They should not automatically be generalized to ChatGPT Enterprise, API usage, or every OpenAI product.
OpenAI published additional adoption data in June 2026.
Its Signals analysis found that users six months after signing up were sending approximately 50% more messages per day than when they first joined and had tried roughly twice as many distinct ChatGPT capabilities. (OpenAI)
The same analysis reported that users predominantly communicating in languages other than English represented more than half of active users.
OpenAI says the 2026 usage-depth analysis was based on a 0.1% sample of qualifying accounts created between October 15, 2025 and May 1, 2026, with activity measured through May 31, 2026.
These are more useful behavioural statistics than an unsourced claim about what "the average ChatGPT user" does because the population and methodology are attached.
If you are evaluating product usage rather than reporting market scale, retained behaviour matters more than downloads alone. Quash's guide to mobile app retention statistics and churn explains why acquisition and continued use should be measured separately.
ChatGPT Web Traffic and Market Position
Web traffic provides another view of ChatGPT's scale, but it is not the same thing as active-user reporting.
Similarweb's generative AI market analysis estimated the following monthly visits to ChatGPT.com during 2025:
Month | Estimated ChatGPT.com visits |
January 2025 | 3.85 billion |
March 2025 | 4.55 billion |
June 2025 | 5.40 billion |
August 2025 | 5.85 billion |
October 2025 | 6.17 billion |
December 2025 | 5.52 billion |
Across the year, Similarweb's published data ranged from approximately 3.85 billion monthly visits in January to a peak of 6.17 billion in October.
The original 3.8–5.85 billion range sometimes quoted for 2025 stops too early in the year and misses ChatGPT's higher September and October traffic.
Similarweb calculated approximately 43% growth in ChatGPT web traffic from January to December 2025. ChatGPT remained the largest generative-AI website in its dataset by visit volume, while competitors such as Gemini grew more rapidly from smaller starting points.
That distinction matters.
A competitor gaining web-traffic share does not prove that ChatGPT's weekly active users, subscribers, mobile usage, or company revenue declined.
Each requires its own measurement.
ChatGPT Mobile App Downloads and Spending
Mobile data provides yet another view of ChatGPT's scale.
Appfigures data reported by TechCrunch estimated that the ChatGPT app had generated $2 billion in global consumer spending from its May 2023 mobile launch through August 2025.
During January through July 2025 alone, Appfigures estimated approximately $1.35 billion in consumer spending, compared with $174 million during the same period of 2024.
That represented a reported 673% year-over-year increase.
Appfigures also estimated:
Approximately 690 million lifetime mobile installs
$2.91 in lifetime global consumer spending per download
Approximately 45 million average monthly downloads at the time of the report
318 million downloads during 2025 through the reporting period
The $2.91 figure is spending per download, not average revenue per active user.
Appfigures also estimated that ChatGPT's cumulative mobile consumer spending was approximately 30 times the combined lifetime spending of Claude, Copilot, and Grok at that point.
These figures describe mobile app-store activity. They do not represent all OpenAI revenue, all ChatGPT subscriptions, or unique people using ChatGPT.
An installation also says nothing by itself about continued activity.
For product teams, acquisition and retention should therefore remain separate measurements.
What These Statistics Do Not Say
Even a carefully sourced statistics page has boundaries.
1. Weekly active users are not website visits
A user may visit ChatGPT.com many times during one week. Similarweb traffic and OpenAI's reported active-user figures measure different things.
2. Downloads are not active users
A mobile installation does not tell you whether the person opened the app once, uses it every day, deleted it, or reinstalled it.
3. Consumer spending is not total OpenAI revenue
Appfigures measures spending through mobile app stores. OpenAI also generates revenue through web subscriptions, business products, API usage, and other channels.
4. OpenAI revenue is not automatically ChatGPT revenue
The reported $40 billion-plus annualized revenue pace applies to OpenAI as a company.
5. Run rate is not completed-year revenue
Annualizing the current revenue pace does not mean the company has already recorded that amount over a full financial year.
6. Different user metrics should not be placed on one growth line
A monthly active user count, weekly active user count, visit estimate, and download total are not interchangeable measures of audience size.
7. Study findings retain the limits of their population
The OpenAI/NBER usage study examined consumer ChatGPT usage. Its work-versus-non-work percentages should not automatically be applied to enterprise or API populations.
A narrower statistic with a clear definition is more useful than a larger number whose denominator has disappeared.
Frequently Asked Questions
How many active users does ChatGPT have in 2026?
ChatGPT reached approximately 1 billion weekly active users in July 2026, according to an OpenAI spokesperson quoted by The Verge.
OpenAI had previously reported more than 900 million weekly active users in February 2026.
Does ChatGPT have 1 billion users?
Yes, with an important qualification.
OpenAI has said more than 1 billion people use ChatGPT, and an OpenAI spokesperson told The Verge that ChatGPT reached approximately 1 billion weekly active users in July 2026.
Always retain the "weekly active users" label when citing that figure.
How many paying ChatGPT subscribers are there?
OpenAI reported more than 50 million consumer subscribers on February 27, 2026. (OpenAI)
How much revenue does ChatGPT make?
There is no verified public figure in this dataset that isolates all ChatGPT revenue.
CNBC reported that OpenAI's company-wide annualized revenue pace exceeded $40 billion in August 2026. That should not be relabeled as ChatGPT-only revenue.
What percentage of ChatGPT use is for work?
OpenAI's consumer-usage study found that approximately 30% of consumer ChatGPT usage was work-related, while approximately 70% was non-work-related.
Those percentages come from the study population and should not be generalized to enterprise usage.
How many downloads does the ChatGPT app have?
Appfigures estimated approximately 690 million lifetime mobile installs through August 2025, according to TechCrunch.
That is an installation estimate, not a count of unique or active users.
How much money has the ChatGPT mobile app made?
Appfigures estimated $2 billion in cumulative global consumer spending through the iOS and Android apps from launch in May 2023 through August 2025.
This is mobile consumer spending, not total ChatGPT or OpenAI revenue.
How much traffic does ChatGPT get?
Similarweb estimated monthly ChatGPT.com traffic between approximately 3.85 billion and 6.17 billion visits during 2025, with the highest published monthly figure in its dataset occurring in October.
Website visits should not be treated as unique users.
Conclusion
The most useful ChatGPT statistics in 2026 are not simply the largest numbers. They are the figures whose scope remains attached.
ChatGPT reached approximately 1 billion weekly active users by July 2026. OpenAI reported more than 50 million consumer subscribers in February. OpenAI's consumer-usage research found a roughly 70% non-work and 30% work-related usage split. Similarweb recorded as many as 6.17 billion estimated monthly ChatGPT.com visits during 2025, while Appfigures estimated 690 million mobile installs and $2 billion in cumulative mobile consumer spending through August 2025.
Separately, OpenAI's company-wide annualized revenue pace exceeded $40 billion by August 2026, according to CNBC.
Each statistic answers a different question.
Weekly active users describe reported reach. Similarweb measures estimated website demand. Appfigures measures mobile app-store activity. Usage research describes behaviour inside the studied population. Revenue run rate describes the current pace of OpenAI's business.
When citing any ChatGPT statistic, keep the date, metric, population, and source attached.
Methodology and Sources
Where possible, first-party OpenAI data is used instead of secondary summaries. Independent estimates are clearly labeled as estimates rather than company disclosures.
Primary and measurement sources include:
Forecasts and statistics without sufficiently inspectable sourcing were excluded.
Continue reading
More articles from the Quash editorial desk.
