Grok AI Statistics & Usage (2026)

You may be looking for one number to answer “how many people use Grok AI?” for a presentation or a market brief. The catch is that Grok’s public figures describe different populations: usage of Grok features, combined Grok/X activity, standalone web visits, app installs, paid subscriptions, and app-market daily activity.

The clearest answer is approximately 117 million monthly active users who used Grok’s AI features as of March 31, 2026, according to SpaceX’s SEC-filed S-1. That disclosure sits within a separate approximately 550 million combined Grok/X MAU measure. You should not add either figure to downloads, visits, or app share; each answers a different question. SpaceX S-1

Key Grok AI statistics for 2026

• 117M Grok-feature MAUs: SpaceX reported approximately 117 million monthly active users who used Grok’s AI features as of March 31, 2026. This is a company disclosure rather than an independent audience estimate.

• 550M combined Grok/X MAUs: The same filing reported approximately 550 million combined monthly active users across Grok and X on March 31, 2026, compared with approximately 520 million on December 31, 2025. Those are combined-platform figures, not Grok-only counts.

• 1.9M active paid Grok-tier subscriptions: As of March 31, 2026, the filing reported approximately 1.9 million active subscriptions across SuperGrok, SuperGrok Heavy, and SuperGrok Lite. Paid subscriptions are not the same measure as monthly active users.

• $818M in AI-segment revenue: For the three months ended March 31, 2026, SpaceX reported $818 million in revenue for its AI segment. The segment includes AI applications and compute infrastructure, so the figure is not standalone Grok revenue. SpaceX S-1

• 15.2% chatbot-app DAU share: Apptopia reported that Grok held 15.2% of daily-active-user share in its generative-AI chatbot app market in January 2026, up from 12% in December 2025. This is app-market DAU share, not global Grok market share. Apptopia’s February 2026 data brief

• 100M+ Android downloads: Google Play displayed 100M+ cumulative downloads for the Grok Android app when the listing was observed on October 1, 2026. This is a store download band, not an active-user count. Grok on Google Play

• 1.5M iOS ratings: The U.S. App Store listing displayed 1.5 million ratings, a 4.9-star rating, and a #7 Productivity position when observed on October 1, 2026. Apple’s listing did not show a cumulative download total. Grok on the App Store

• 202.7M standalone web visits in March 2025: Similarweb’s published table recorded 202,650,131 visits to grok.com in March 2025, after 1,205,770 in January and 51,462,811 in February. Visits are sessions to the standalone site, not total Grok users across X and mobile apps. Similarweb

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.

Email addressGet Instant Access

100% Free. No spam. Unsubscribe anytime.

How many people use Grok AI?

The best current company-disclosed answer is about 117 million Grok-feature MAUs as of March 31, 2026. In the S-1, a monthly active user is someone who interacted with Grok or X through a web browser or mobile app at least once during the relevant 30-day period.

That definition makes the number useful, but specific. It measures people using Grok’s AI features within an ecosystem that includes X; it does not establish a count of people who use only the standalone Grok website or app. The filing says the combined measure seeks to account for people who access both Grok and X, so it should not be treated as the sum of separate service audiences. Read the MAU definitions and disclosures in the S-1

The 520M combined Grok/X MAUs on December 31, 2025 and 550M on March 31, 2026 do show growth in the broader combined platform measure. They do not create a verified Grok-only growth series. The filing does not provide a Grok-specific December 2025 MAU in the disclosed figures, so a quarter-over-quarter Grok-only rate would be unsupported.

Paid Grok adoption and AI-segment financials

The closest public measure of paid Grok adoption is the approximately 1.9 million active subscriptions across SuperGrok, SuperGrok Heavy, and SuperGrok Lite as of March 31, 2026. This is a count of active paid subscriptions across named tiers, not a count of distinct users and not a conversion rate from the 117M Grok-feature MAUs.

SpaceX also reported approximately 6.3 million active paid subscribers across the AI segment at that date. That segment total included approximately 4.4 million X Premium and Premium+ subscribers, alongside the Grok-branded tiers. If you are comparing paid adoption with usage, preserve that distinction: the 1.9M figure is the narrower Grok-tier subscription measure, while 6.3M is the broader AI-segment subscription measure.

For the three months ended March 31, 2026, the filing reported $818 million of AI-segment revenue and a $2.469 billion AI-segment loss from operations. Those financial figures cover AI applications and compute infrastructure. They show the scale of the reported segment, but they do not reveal standalone Grok revenue. SpaceX’s segment reporting

Grok website traffic and app adoption

Standalone grok.com traffic rose quickly during the product’s early 2025 period. Similarweb published the following visit counts:

These are traffic observations, not unique-person counts. They also omit activity through X and native mobile apps, so they cannot be combined with the company’s MAU disclosure. Similarweb described the surge as part of a broader generative-AI traffic shift; it is evidence of web demand, not a retention measure. Similarweb’s published traffic table

The Android and iOS stores offer different signals. Google Play’s 100M+ figure is cumulative downloads, which can include reinstallations and people who later stop using the app. The App Store’s 1.5M ratings, 4.9 rating, and #7 Productivity position show store reception and chart standing, but neither ratings nor ranking is a user total.

If you need mobile-platform context for interpreting store measures, Quash’s guide to iOS and Android market-share statistics explains why platform-level app signals and actual active usage are different measurements. That distinction matters especially for Grok because its reported usage also spans X and the web.

What does Grok’s 15.2% app-market share mean?

Apptopia reported 15.2% daily-active-user share for Grok in the generative-AI chatbot app market in January 2026, rising from 12% in December 2025. The population here is the app market tracked by Apptopia, and the metric is its share of daily active users within that market.

You should not read 15.2% as Grok’s share of all global AI use, all chatbot use, or all activity across Grok’s web, mobile, and X surfaces. It is an app-market measurement from a mobile-app intelligence provider. Used carefully, it is a useful indicator of Grok’s competitive position in that defined mobile category. Apptopia’s January 2026 DAU-share analysis

Third-party estimates of Grok usage

Company disclosures are not the only numbers available. Presenc AI estimates approximately 95 million Grok MAUs in Q1 2026, compared with approximately 35 million in Q1 2025. It also estimates that 78% of Grok usage occurred through X integration.

These figures should remain explicitly labelled as Presenc AI estimates. The published material does not fully disclose the methodology needed to reconcile its 95M estimate with the SEC filing’s 117M company-defined Grok-feature MAUs. The difference is not a reason to average the figures or assume either source is wrong; they may cover different definitions, collection methods, or inclusion rules. Presenc AI’s Grok usage estimates

For a citation-ready usage statement, prefer the SEC’s dated company disclosure. Use the Presenc estimate only when you need a clearly labelled third-party view rather than a company-reported measure.

Grok model statistics are not usage statistics

Model capabilities can help explain product context, but they should not appear in a total of users, visits, or downloads.

In its September 19, 2025 announcement, xAI said Grok 4 Fast had a 2 million-token context window. That was a dated product-availability statement, not evidence of adoption or active use. xAI’s Grok 4 Fast announcement

xAI’s August 12, 2026 announcement reported a score of 61 for Grok 4.6 on the Artificial Analysis Intelligence Index, which xAI described as a composite of nine benchmarks. Artificial Analysis also discussed Grok 4.6’s benchmark results, but published scores can depend on model variant, evaluation configuration, and snapshot date. Treat the 61 as dated model-performance context, not as an audience or market-share statistic. xAI’s Grok 4.6 announcement Artificial Analysis’s Grok 4.6 analysis

Why Grok statistics do not add up

The apparent disagreement in Grok statistics usually comes from category errors rather than competing answers to one question.

• Grok-feature MAU is a company-defined 30-day usage measure.

• Combined Grok/X MAU is a broader, deduplicated cross-platform measure.

• DAU share is a share of daily activity in a defined chatbot-app market.

• Web visits are sessions on grok.com, not necessarily unique people.

• Downloads are cumulative installations or a store-provided install band.

• Ratings and chart positions are App Store signals, not usage counts.

• Paid subscriptions count active paid subscriptions, not all monthly users.

• AI-segment revenue covers a reported business segment rather than Grok alone.

• Benchmark scores evaluate model performance, not adoption.

Keeping those labels beside the number is more useful than forcing a single “total Grok users” figure. If you are compiling AI adoption data more broadly, the same discipline applies to AI testing statistics and adoption data: population, period, and measurement method determine whether two figures can be compared.

What is still unknown about Grok usage?

Public disclosures and named data sources do not establish standalone Grok revenue, one global Grok market-share percentage, global query volume, retention, churn, or a verified Grok-only MAU for December 2025. They also do not support adding MAUs, downloads, visits, ratings, and DAU share into a single total.

No first-party Quash data covers Grok adoption, Grok usage, Grok-specific QA patterns, customer language, or a completed Grok experiment. There is therefore no Quash telemetry to use as evidence in this report.

Those gaps matter when you are making a business case. The public record can support a claim that Grok has meaningful measured reach across feature usage, paid tiers, web traffic, mobile distribution, and app activity. It cannot support a precise, all-channel count of unique Grok users.

Methodology and source notes

This report prioritizes sources that own the underlying measure: SpaceX’s SEC filing for company disclosures, Similarweb for its traffic dataset, Apptopia for its app-market data, Google Play and Apple for store signals, and xAI for product announcements. Presenc AI is included as a named third-party estimator, not as a substitute for the company disclosure.

Store data is particularly time-sensitive. The Google Play download band and App Store rating and rank cited here were observations from October 1, 2026. The cited web series covers January through March 2025, while the app-share measure covers January 2026. Keeping the observation period beside each number prevents a launch-period traffic spike, a current store listing, and a quarterly company disclosure from being presented as though they describe the same moment.

The most defensible way to use these Grok AI statistics is to select the measurement that matches your question. Use 117M for company-disclosed Grok-feature MAUs, 1.9M for disclosed active subscriptions across Grok-branded paid tiers, 15.2% for Apptopia’s January 2026 chatbot-app DAU share, and the traffic or store figures only for their narrower channels. Your decision should follow the measure—not the biggest number.