X (Twitter) shadowban check: how to test it in 2026
X’s help center says “we don’t shadow ban”, and its other pages describe each limit people mean by the word. We read them and X’s open-source code, tested 14 free checkers on 12 October 2026, and wrote down the manual checks you can run from a second account.
To check for a shadowban on X in 2026, start with X’s own Under the Hood report at x.com/i/under_the_hood. It lists the labels X put on your posts and your account last month, but none of them names search or the order of replies in a thread. Those you check by hand, from a second account that doesn’t follow you, because X’s search no longer works when you’re logged out.
Free shadowban checkers give a rough first look at best. Of the 14 we tested on 12 October 2026, one said an account that does not exist “exists” and has “No Search Ban”. The manual checks are in the next section.
Does X shadowban accounts?
X says it doesn’t. Its help page “Debunking X myths” answers: “Simply put, we don’t shadow ban! Ever.” Twitter said the same in its 2018 post. But X does limit who sees some posts and accounts. Since 2023 it has called this restricting reach, “also known as visibility filtering”, and admitted that “we have not historically been transparent when we’ve taken this action” (X Safety). Its enforcement page lists what that can mean for a post:
Whole accounts can be filtered too: X “may automatically remove accounts engaging in these behaviors from search” (search rules). The names people use for these limits come from checkers, not from X:
| Community name | What people mean | Nearest wording at X |
|---|---|---|
| Search suggestion ban | Your account is missing from the dropdown when someone types your name. | No current term. Account search prefers profiles with “a complete name, username, and bio”. |
| Search ban | Your posts don’t come up in search, even for from:yourhandle. | “X may automatically remove accounts engaging in these behaviors from search”. |
| Ghost ban (thread ban) | Your reply is missing from the thread when other people open it. | “Restricting the post’s discoverability to the author’s profile”. |
| Reply deboosting | Your reply sits low in the thread or behind a button. | “Downranking the post in replies”, and the “Show probable spam” button. |
| Kept out of recommendations | Followers see your posts; people who don’t follow you rarely do. | Under the Hood labels: “hidden from recommendations to non-followers”. |
Sources: X’s pages on search and replies, its enforcement page and the Under the Hood label file. Locks, captchas and limited states are a separate matter: X tells you about those with a message when you log in.
How do you check if you’re shadowbanned on X yourself?
Use a second account that doesn’t follow you and hasn’t interacted with you, logged in. The old advice to open a private window no longer works: on 12 October 2026 a logged-out search sent us to the sign-in screen, and X says “Advanced search is available when you’re logged in to X.com” (advanced search page).
A single post still opens logged out, but only the top of its thread loads, so it can’t tell you whether your reply is further down.
- 1
Reply count. The post had 196 replies.
- 2
The first replies. X loaded three of the 196. Two are in full view, and the third fades under the prompt. The names and pictures of the people replying are blurred by us.
- 3
Sign-in prompt. “See all the replies” and “Continue to X” cover the rest of the thread.
A post by @NASA of 8 October 2026, opened logged out on 12 October 2026 and captured at 2x.
From the second account:
Search from:yourhandle — Open the Latest tab. If an account that has posted gets no results, checkers call that a search ban.
Type @yourhandle — Type it into the search box without pressing Enter. If you’re missing from the dropdown, that’s a search suggestion ban.
Search "a phrase from your post" — Pick one of your recent posts and search its own words, the way a stranger would find it.
Open three threads you replied in — Pick posts by other accounts from the last few days, ideally with plenty of replies, and scroll each one for your reply.
Look behind “Show probable spam” — If your reply shows only after that button, X put it with the replies it rates as potentially lower quality and spam.
Look for a pattern — One low reply proves little: X ranks a reply higher when the author replied, when the viewer follows you, or when you have Premium. The signal is a reply missing or folded away under several posts.
Before reading a result as a ban, rule out what X lists on its page about search not working: protected posts, an unconfirmed email address, potentially sensitive content, a profile without a name or bio. Ranking factors and the button are from X’s page on replies. We had no restricted account to run these steps on, so we can’t show what a positive result looks like.
What does X’s Under the Hood report show?
It lists the labels X’s systems applied to your posts and account last month and what each does to visibility. Since 19 September 2026 it is open to accounts with “10+ posts in prior month, account at least one year old” (@XOpenSource), and the report page’s code offers “Download this report (JSON)”. A report needs a login, so we have no screenshot; the labels below are from X’s open-source code.
| Label | What X says it means | Effect on visibility |
|---|---|---|
| SPAM_HIGH_RECALL (post) | “Post detected by automated systems as one that may contain spam.” | “Post hidden from recommendations to non-followers.” |
| SPAM (post) | Identified as likely to break X’s Authenticity policy. | “Post is not shown on X.” |
| FOSNR_ABUSE, FOSNR_HATEFUL_CONDUCT and others (post) | Reported or detected under the abuse, hateful conduct, violent content or civic integrity policies. | “Post discoverability is restricted to the author’s profile.” A label tells all users. |
| SpamHighRecall (account) | “Account detected by automated systems as likely to post spam.” | “The account’s posts are hidden from recommendations to non-followers.” |
| DoNotAmplify (account) | Detected as likely to break X’s Terms of Service. | Posts hidden from recommendations to non-followers “pending further review.” |
The label file lists more, mostly with the same effect. None of them names search or the order of replies, so a clean report doesn’t rule out what the manual checks look for.
Do shadowban checkers still work in 2026?
Some return results, and none publishes how accurate it is. On 12 October 2026 we ran 14 free checkers on @XDevelopers, X’s developer account, then the ten that answered on @qz8x7nx2026ab, a handle with no account behind it.
| Checker | @XDevelopers | A handle with no account |
|---|---|---|
| Shadowban Test (yuzurisa) | No ban of any of the four kinds | “does not exist or doesn’t have any tweets” |
| HiSubway | No ban of any of the four kinds | “@qz8x7nx2026ab exists.” and no ban of any of the four kinds |
| xshadowban.com | “No restrictions detected”, “High confidence” | “Likely shadowbanned”, “High confidence” |
| Circleboom | “No issues found” | “The account does not exist!” |
| OpenTweet | “Could not determine search visibility” | “Account Suspended”, then “suspended or does not exist” |
Of the other nine, ShadowBird, ShadowCheck, Shadowban Radar and NotPeople said the handle had no account, Sorsa’s playground returned an error, and Postory, SocialCal, Volumn.ai and Postel stopped on @XDevelopers at a captcha, a rate limit, a missing API key or no result at all.
Checkers test your account from outside and can only report a ban they find. The Shadowban Test prints on every result: “We can only detect the presence of specific bans or deboosting methods, not their absence.” We had no restricted account to test them on, so we can’t say how often any of them catches a real one.
Replies every day from your own account, in your voice, with a daily ceiling on every plan.
Book a callWhat gets your reach limited on X?
X names the behaviour itself, in its search rules, authenticity policy and automation rules:
WHAT X’S RULES NAME
“Repeatedly post duplicate or near-duplicate content (links or posts).”
“Sending bulk, aggressive, high-volume unsolicited replies, mentions, or direct messages”.
“promoting content by replying with content that is irrelevant to the topic of the original post”.
“posting and deleting the same content repeatedly”.
“Aggressively follow and unfollow people.”
A rule in X’s code labels a post whose link fails X’s URL check SEARCH_BLACKLIST and DO_NOT_AMPLIFY, among others, and a study of more than 40 million posts (NDSS 2026) found links reduced visibility “by up to a factor of eight”. Volume counts too, though X publishes no threshold; see how many replies per day on X.
What should you do if your reach is limited?
If X showed you a message, follow it first: each names its own way out. X publishes no duration; for accounts limited after aggressive following or engagement, its help page says only “for a specified time, which varies based on the nature of the violation.” Guides give their own estimates, from 12–48 hours (Tweet Archivist, search suggestion ban) to up to 30 days (Sorsa, ghost ban).
Download Under the Hood If your account qualifies, see which labels X applied last month.
Run the manual checks From a second account: search, the dropdown, and three threads you replied in.
Stop what X calls spam Repeated or near-identical replies, off-topic replies with links, posting and deleting the same thing. “You are ultimately responsible for the actions taken with your account, or by applications associated with your account.”
Appeal Labels on posts can be appealed from the label itself. If your account was limited by mistake, X says to contact its support team.
X’s filters also get it wrong. In March 2026 Nikita Bier, then head of product at X, wrote that a new spam filter had “falsely tagged subset of accounts” and that “99% of those suspensions were reverted”. If an app or a service posts on your account, ask what it posted before your reach dropped.
How GhostReply sets the pace on your account
With GhostReply, replies go out every day from your own X account, in your voice, under posts your market already reads. Access goes through X’s delegate feature, so no password changes hands; how that works is in our guide to X’s delegate feature. Every plan has a ceiling of 50 replies a day, and below it you choose the pace.
Nobody outside X can promise that an account will never be limited, and we don’t. If a label, a shadowban or a fall in reach shows up, our terms say what happens next: “We lower the volume and carry on.” If you’d rather run the routine yourself, the reply guy strategy, step by step covers it.
Frequently asked questions
How do I know if I’m shadowbanned on X?
Check from a second account that doesn’t follow you: search from:yourhandle on the Latest tab, type your handle into the search box, and look for your replies in a few threads. If your account is over a year old with 10 or more posts last month, X’s Under the Hood report also lists the labels on your posts and account.
Does X tell you if you’re shadowbanned?
Partly. Since 19 September 2026, accounts at least a year old with 10 or more posts in the previous month can open X’s Under the Hood report at x.com/i/under_the_hood. It lists last month’s labels and what each one does, mostly keeping posts out of recommendations to non-followers. None of the label descriptions names search or the order of replies.
Can I check for a shadowban in incognito mode?
Not any more. On 12 October 2026 a logged-out search on x.com sent us to the sign-in screen, and X’s help page says advanced search is available when you’re logged in. A post opened logged out loaded only its first three replies. Run the checks from a second account that doesn’t follow you.
Are shadowban checker sites accurate?
None of the 14 we tried publishes its accuracy. On 12 October 2026, one reported an account that doesn’t exist as existing with no bans, another rated it “Likely shadowbanned” with high confidence, and a third headlined it as suspended. The Shadowban Test at yuzurisa.com prints on every result: “We can only detect the presence of specific bans or deboosting methods, not their absence.”
Why can’t other people see my replies on X?
X’s help pages give three reasons. Replies rank higher when the post’s author replied, when the viewer follows you, or when you have Premium. Replies X rates as likely spam go behind a “Show probable spam” button. And in timelines, a back-and-forth between two people shows only to those who follow them.
How long does a shadowban on X last?
X publishes no duration. Its help page says limited features last “for a specified time, which varies based on the nature of the violation.” Guides that rank for the question give anything from 12 hours to 30 days, and none of the four we checked names a source for its figures.
Do links get you shadowbanned on X?
X’s rules count links shared without commentary as spam when they make up most of your activity, and a rule in its open-source code labels posts with unsafe links SEARCH_BLACKLIST. A study of more than 40 million posts, published at NDSS 2026, found that posts with external links got up to eight times less visibility.
You choose the pace, below your plan’s daily ceiling.
Daily replies from your own account, in your voice, under posts your market already reads.
Up to 50 replies a day on every plan. If a label or a drop in reach shows up, we lower the volume.
You approve examples of your voice and the accounts to reply under before the start.
Access through X’s delegate settings, no password. Remove us on the same screen at any moment.