How Many People Use X (Twitter)? User Statistics

TL;DR: About 550 million people used X or Grok each month at March 31, 2026, according to the audited SpaceX prospectus. X has not published an audited X-only number since it went private in 2022. In the EU, X counted 64.7 million monthly active recipients in the first half of 2026, and 21% of US adults use it. Our own sample found X search showing at least 112 million new posts a day in just 12 languages, not counting reposts.

Last updated October 9, 2026. Every third-party figure below links to its source, and each figure carries the period it describes. Our own measurements were taken with the GetXAPI API on October 8 and 9, 2026; the method is at the end.

Most pages that answer how many people use Twitter, now called X, repeat the same four numbers without dates: a 600 million claim, a 2013 posts-per-day figure, advertising reach presented as users, and a statistics-site estimate. This guide separates what was audited from what was claimed, and adds data we measured ourselves.

How many people use X (Twitter)?

About 550 million people use X or Grok each month. That figure comes from the audited prospectus SpaceX filed with the US Securities and Exchange Commission in June 2026, which reports roughly 550 million monthly active users at March 31, 2026 and 520 million at December 31, 2025, counted across X and Grok together and deduplicated by sign-in (SpaceX 424B4, SEC). It is the only audited usage number published since Twitter went private in October 2022. About 117 million of those users used Grok's AI features, so X on its own is at most 550 million and likely lower.

Three other figures help frame it:

Measure	Figure	Period	Source
Monthly users, X and Grok combined (audited)	about 550 million	March 31, 2026	SpaceX prospectus
Monthly active recipients in the EU	64,727,103	January to June 2026	X DSA transparency report
Share of US adults who use X	21%	February to June 2025 survey	Pew Research Center
Daily users of the X mobile app (estimate)	about 125 million	January 7, 2026	Similarweb via TechCrunch

If you need one number for a slide, use "about 550 million monthly users across X and Grok (March 2026)" and keep the Grok qualifier. Quoting 600 million, or anything higher, means quoting an unaudited claim.

Why the numbers disagree: six ways X users get counted

Six ways X (Twitter) users get counted. Only audited filings define and check the number; ad reach counts accounts, not people.

Search for X user numbers and you will find anything from 125 million to 650 million. Almost none of that spread is error. It comes from mixing six different kinds of count, each measuring something else.

Method	What it counts	Latest figure	Reliability
Company claim	Whatever the company chooses to define as a user	600 million monthly (May 2024)	Low: no definition, no audit
Audited filing	Monthly or daily active users, defined and audited	about 550 million monthly, X plus Grok (March 2026)	Highest
EU regulator report	Average monthly active recipients in the EU, required by the Digital Services Act	64.7 million (H1 2026)	High, but EU only
Ad reach	Accounts an advertiser can target, including inactive and duplicate accounts	586 million (January 2025)	Not a user count
App panel	Estimated daily users of the mobile app from a measured panel	about 125 million (January 2026)	Medium: app only, modelled
Survey	Share of adults who say they use X	21% of US adults (2025)	High for one country

The company claim comes from Elon Musk's posts: "over 540 million" monthly users in July 2023 (Musk on X, figure reported by Reuters via Al Jazeera) and 600 million in May 2024 (Social Media Today). Ad reach comes from DataReportal, which states plainly that its 586 million is advertising audience, not monthly active users. Several statistics sites present that same number as "users". It is not.

Two rules make the rest of this guide easy to read. First, never compare a monthly figure with a daily one. Second, check whether a number is a claim or an audited count before you quote it.

Twitter monthly active users over time

While Twitter was a public company it reported monthly active users (MAU) every quarter, defined and checked by auditors. These figures come from its SEC filings, at year end where available.

Period	Monthly active users	Source
Q2 2012	151.4 million	S-1, 2013
Q4 2013	240.9 million	10-K FY2013
Q4 2014	288 million (later restated to 292 million)	10-K FY2014
Q4 2015	320 million (later restated to 305 million)	10-K FY2015
Q4 2016	319 million	10-K FY2016
Q4 2017	330 million	10-K FY2017
Q4 2018	321 million	10-K FY2018
Q1 2019	330 million (the last MAU figure Twitter reported)	10-Q Q1 2019

After the first quarter of 2019 Twitter stopped reporting MAU and switched to a narrower daily metric (next section). The restatements matter: Twitter removed users who only followed accounts by SMS, which cut the Q4 2015 figure from 320 million to 305 million. Whichever number you quote, quote it from the filing you cite.

Twitter reported audited user numbers every quarter until it went private in 2022. Since then X has published claims and one audited figure that includes Grok.

Since the 2022 takeover the series has become scattered claims rather than reports: over 540 million monthly users (Musk, July 2023), 600 million (Musk, May 2024), then the audited 520 million (December 2025) and about 550 million (March 2026) in the SpaceX prospectus. The audited figures include Grok, so they are not directly comparable with Twitter's old X-only MAU. What they do show is that the combined X and Grok audience in 2026 is below Musk's 2024 claim for X alone.

How many people use X every day?

Twitter's last official daily figure was 237.8 million monetizable daily active users (mDAU) in the second quarter of 2022 (Q2 2022 shareholder letter). mDAU counted logged-in people who could see ads on a given day, so it was always smaller than total daily use.

Period	mDAU	Of which US	Source
Q4 2018	126 million		10-K FY2018
Q4 2019	152 million	31 million	10-K FY2019
Q4 2020	192 million	37 million	10-K FY2020
Q4 2021	217 million	38 million	10-K FY2021
Q2 2022	237.8 million	41.5 million	Q2 2022 letter

Since then there is no audited daily number. The SpaceX prospectus reports monthly users only. Independent estimates come from app panels: Similarweb measured about 132 million daily users of the X mobile app in June 2025, down 15.2% from a year earlier (TechCrunch), and about 125 million on January 7, 2026 (TechCrunch). Those counts leave out people who only use x.com in a browser, and you cannot add web visits to app users, because the same people appear in both.

Is X gaining or losing users?

On the evidence available, X has been flat or shrinking since 2024, with the caveat that no audited X-only figure exists to settle it.

Signal	What it shows	Source
X's own numbers	600 million monthly users claimed in May 2024; about 550 million audited in March 2026, and that includes Grok	Social Media Today, SpaceX prospectus
App usage	Mobile daily users down 15.2% year on year in June 2025	Similarweb via TechCrunch
Competition	Threads passed X in daily mobile users in January 2026 (141.5 million against 125 million)	Similarweb via TechCrunch
EU	64.8 million monthly recipients in H2 2025 and 64.7 million in H1 2026, flat across the two latest reports	X DSA reports
US	22% of adults in 2023, 21% in 2024 and 2025	Pew Research Center

Two cautions. The website still draws far more daily traffic than Threads does on the web (145.4 million daily web visits for X against 8.5 million for Threads in January 2026, per the same Similarweb data), so "Threads overtook X" is only true on mobile. And earlier EU reports showed higher figures, which may partly reflect a change in how X counts recipients, so do not read the EU series as a straight decline.

X users by country

There is no official per-country user count for the whole world. Two sources come close, and they measure different things.

EU countries, official (H1 2026). X's Digital Services Act report breaks its 64.7 million EU recipients down by member state (X DSA transparency report):

EU country	Monthly active recipients
France	12.39 million
Spain	11.03 million
Germany	9.95 million
Netherlands	5.07 million
Italy	5.07 million

Worldwide, advertising reach (January 2025). DataReportal reports how many accounts advertisers can reach in each country (DataReportal). This counts accounts, not people, and includes inactive ones:

Country	Ad reach
United States	104 million
Japan	70.9 million
Indonesia	25.2 million
India	24.1 million
United Kingdom	22.9 million

Japan's place near the top shows up in our own measurement too. In our sample, Japanese posts arrived at almost the same rate as English posts in several windows.

Who uses X? US demographics

Pew Research Center's 2025 survey of US adults, conducted from February to June 2025, is the most careful demographic source (Pew fact sheet).

Group	Share who use X
All US adults	21%
Ages 18 to 29	33%
Ages 30 to 49	25%
Ages 50 to 64	16%
Ages 65 and over	10%
Men	25%
Women	16%

X skews young and male in the US. Pew changed its survey method in 2023, so compare 2023 onward with each other, not with earlier years.

How many posts are sent on X per day?

The famous figure of 500 million posts a day dates from 2013. Twitter stated "approximately 500 million Tweets every day" in its 2013 IPO filing (S-1). In October 2023 X's then chief executive Linda Yaccarino repeated 500 million and broke it down as about 100 million original posts, 100 million replies and 300 million reposts and quotes (TechCrunch). X has published nothing newer.

Since nobody publishes a current number, we measured one. We counted every post X search returned in fixed five-second windows, once every four hours across a full day, for 12 languages. Search does not return reposts, so this counts original posts, replies and quotes.

Window (UTC)	Posts per second, 12 languages
Oct 8, 07	1,379
Oct 8, 11	1,549
Oct 8, 15	1,736
Oct 8, 19	774
Oct 8, 23	1,423
Oct 9, 03	916
Average	1,296, about 112 million a day

That is a floor, not a total. English and Japanese reached our page limit in most windows, so their true rates are higher, and the 12 languages leave out many others. Even so, it is the same order as Yaccarino's 2023 split of about 200 million originals and replies a day. Average rates by language across the six windows:

Language	Posts per second (at least)
English	422
Japanese	332
Spanish	191
Arabic	78
Portuguese	75
Korean	42
Indonesian	41
Turkish	39
French	30
Thai	28
German	9
Hindi	9

How many posts per day can one account send? X's help centre lists 50 original posts and 200 replies per day for unverified accounts, enforced in semi-hourly chunks, along with 500 direct messages and 400 follows a day (X limits). Before May 2026 the posting limit was 2,400 updates a day for everyone (heise online).

What a typical X post and poster look like

We sampled the 300 newest posts in each of 12 languages on October 9, 2026 (3,600 posts from 3,490 different accounts) and read the post and author fields the API returns.

Measure	Value
Posts that are replies	64.5%
Posts that quote another post	10.4%
Posts with an image or video	15.2%
Posts with a link	16.6%
Median follower count of the poster	280
Posters with fewer than 100 followers	30.6%
Posters with fewer than 1,000 followers	69.7%
Posters with 10,000 or more followers	6.5%
Posters with a blue check	18.8%
Posters whose account was created in 2025 or 2026	23.4%
Posters whose account predates 2015	16.8%
Median account age	4.6 years

X is mostly a conversation between small accounts. Two thirds of new posts are replies, and seven in ten posters have fewer than 1,000 followers. Almost a quarter of active posters joined in the last two years, so the active base keeps turning over even if total usage is flat.

Reply habits differ sharply by language: 82% of German posts and 77% of French posts were replies, against 48% of Japanese and 40% of Thai posts, where original posting dominates. Blue checks are least common among Spanish (6.1%) and Portuguese (5.7%) posters and most common among Thai (31.8%), Hindi (30.9%) and English (27.3%) posters.

At the other end of the scale, the most followed accounts are enormous by comparison. On October 8, 2026 at 17

UTC, @elonmusk had 241.7 million followers, followed by @BarackObama (118.9 million), @Cristiano (114.6 million), @realDonaldTrump (111.9 million) and @narendramodi (107.2 million).

What share of X accounts are bots?

There is no reliable current figure for the percentage of bots on X. The two numbers usually quoted are old and measure different things:

Estimate	What it measured	Source
Fewer than 5%	False or spam accounts as a share of monetizable daily users, Q2 2022	Twitter 10-Q
9% to 15%	Bots as a share of active accounts, from a 2017 academic study	Varol et al., ICWSM 2017

No peer-reviewed study since has measured the whole platform. What we can measure is narrower and still telling. We pulled the 1,000 newest followers of each of the five most followed accounts (5,000 accounts in total) on October 9, 2026:

Newest followers of	Never posted	Created in the last 30 days	No posts, no bio and no followers
@elonmusk	46.8%	36.0%	38.6%
@BarackObama	55.0%	55.6%	21.8%
@Cristiano	77.8%	89.1%	70.1%
@realDonaldTrump	80.0%	74.2%	70.0%
@narendramodi	60.2%	66.8%	47.7%
All 5,000	64.0%	64.3%	49.6%

Half of these new followers are completely empty profiles. That is not proof they are bots: most were created in the last 30 days, so many are likely people who joined recently and have not posted yet. It does mean that follower counts on very large accounts grow partly through accounts that never become active. If you need to judge a specific account, our Twitter bot detection guide covers the signals, and the free fake follower checker runs an audit for you.

How to track X follower counts over time

X's built-in analytics only covers your own account. To build a follower count history for any account, look the account up once a day and store the result. This script does exactly that with GetXAPI's user/info endpoint:

import csv
import datetime
import os

import requests

API_KEY = os.environ["GETXAPI_KEY"]
ACCOUNTS = ["NASA", "elonmusk", "your_brand"]


def follower_count(username: str) -> int:
    r = requests.get(
        "https://api.getxapi.com/twitter/user/info",
        params={"userName": username},
        headers={"Authorization": f"Bearer {API_KEY}"},
        timeout=30,
    )
    r.raise_for_status()
    return r.json()["data"]["followers"]


today = datetime.date.today().isoformat()
with open("followers.csv", "a", newline="") as f:
    writer = csv.writer(f)
    for name in ACCOUNTS:
        writer.writerow([today, name, follower_count(name)])

Run it once a day with cron:

0 9 * * * /usr/bin/python3 /path/to/track_followers.py

Each lookup costs $0.001, so tracking 10 accounts daily costs about $0.30 a month and 100 accounts about $3. The same response also returns following, statusesCount (total posts) and createdAt, so you can chart posting activity alongside followers. To pull the follower lists themselves rather than the count, see the Twitter followers API.

How to reproduce our posts-per-second count

The counting method is simple enough to rerun for any language or hour. advanced_search accepts since_time and until_time operators in Unix seconds, so you can page through every post in a short window and count unique IDs:

import os

import requests

API_KEY = os.environ["GETXAPI_KEY"]


def count_posts(lang: str, start: int, seconds: int = 5) -> int:
    query = f"lang:{lang} since_time:{start} until_time:{start + seconds}"
    ids, cursor = set(), None
    while True:
        params = {"q": query, "product": "Latest"}
        if cursor:
            params["cursor"] = cursor
        r = requests.get(
            "https://api.getxapi.com/twitter/tweet/advanced_search",
            params=params,
            headers={"Authorization": f"Bearer {API_KEY}"},
            timeout=60,
        )
        r.raise_for_status()
        page = r.json()
        ids.update(t["id"] for t in page["tweets"])
        cursor = page.get("next_cursor")
        if not page.get("has_more") or not cursor or not page["tweets"]:
            return len(ids)

Each page returns about 20 posts for $0.001. A five-second English window at a busy hour can run past 100 pages, so set a cap, and treat a capped window as a minimum, as we did above.

Method and limits

• Posts sample: advanced_search with product=Latest and a language-only query (lang:en, lang:ja and so on) for 12 languages, 300 newest posts each, collected on October 9, 2026 between 04
  and 04
  UTC. Likes and views were not analysed, because posts seconds old have not collected any yet.
• Posting rate: every post in six five-second windows four hours apart (October 8, 07
  UTC to October 9, 03
  UTC), 12 languages, deduplicated by post ID. English and Japanese hit a 120-page cap in most windows, so their figures are minimums. Search results exclude reposts and any posts deleted before we counted.
• Follower audit: the 1,000 newest followers of each of the five most followed accounts, pulled on October 9, 2026, with "empty" meaning zero posts, no bio and zero followers at that moment.
• Top accounts: user/info lookups on October 8, 2026 at 17
  UTC.
• Limits: one day of data, a fixed set of 12 languages, and only what X search exposes. These are snapshots, not a census. Third-party figures are quoted with their own period and source, and none of them is our estimate.

Get X data for your own research

Every number we measured here came from two endpoints and cost a few dollars in total. If you want follower histories, posting rates, audience samples or your own statistics on X, create a free GetXAPI account: there is no X developer account to apply for, standard calls cost $0.001, and the full endpoint reference is in the documentation.