Telegram Spam Restriction: 5 Steps to Get It Lifted

The message arrives without warning and it is always the same wording: you can only send messages to mutual contacts. Nothing was announced, no email came, and the account still opens normally. It just cannot reach anyone new. Most operators respond by doing the two things that make it permanent, which is testing whether it is really enforced and then starting again from a fresh account with the identical behaviour. This guide covers what the restriction actually is, how to tell which one you have, and the five steps that give an account its best chance of coming back. It also covers when an account is not worth saving, because that call is usually made too late.

What the Restriction Actually Is

Telegram's anti-spam system does not ban outreach accounts so much as narrow them. The account stays alive, keeps its history, keeps its groups, and keeps working with people it already knows. What it loses is the ability to initiate contact with strangers, which happens to be the only thing an outreach account exists to do.

The Wording You Will See

Two phrasings cover almost every case. The first says you can only send messages to mutual contacts, which appears when you try to message someone who does not have you saved. The second is the API level PEER_FLOOD, which is the same restriction seen from the other side of the wire. If you run automation you will meet the second first, because the code arrives before anyone reads a screen.

There is a third form worth knowing, which is silence. Some restricted accounts return a success response to the client and simply never deliver, so the campaign dashboard shows sends climbing while nothing arrives. If your delivery numbers look healthy and your reply rate has fallen to zero overnight, check a send against an account you control before you conclude the offer stopped working.

What Still Works

Messages to existing contacts still go through. Groups and channels you already belong to behave normally. Replies to people who messaged you first work. This is exactly why the restriction is so often missed: the account looks fine in ordinary use and only fails on the specific operation you care about.

This partial function is also why the restriction survives so long undetected in a team. The person running outreach sees failures; the person who owns the account opens it, sends a message to a colleague, and reports that it works fine. Both are telling the truth about different operations. Any account health check that consists of opening the app and looking around will pass a restricted account every time.

It Is Not a Ban

A ban removes the account. A restriction leaves it in place and blocks one class of action. The distinction matters because the recovery paths are completely different, and because a restricted account still holds value that a banned one does not. Treat the restriction as a suspended licence rather than a confiscated car.

The practical consequence is that you should not delete a restricted account, log out of it, or reinstall over it. Those actions destroy the session and sometimes the recovery path, and none of them lift anything. Whatever you decide about the account later, the correct immediate state is untouched.

How to Tell Which Restriction You Have

Before doing anything, work out what you are actually dealing with. Three states look similar from the outside and need three different responses, and guessing wrong costs either an account or a week.

Spam Restriction

Applied deliberately by the anti-spam system, usually after reports. New contact is blocked, the state persists across sessions and devices, and it does not clear on its own schedule that you can predict. This is the one that needs the appeal.

The tell is consistency. It follows the account rather than the device, it survives a restart, and it applies to strangers specifically. If you hand the same session to another machine on another network and the behaviour is identical, you are looking at an account level judgement rather than anything local.

Temporary Cooldown

A rate limit rather than a judgement. It appears after sending quickly, it clears by itself, and it usually announces its own duration if you are working through the API. Nothing needs to be appealed here; the account needs to slow down, not apologise.

Cooldowns are also the most useful diagnostic you get, because they arrive with a number attached. Logged across a run they draw a curve, and the shape of that curve tells you whether your pacing is sustainable long before anything is restricted. Operators who record them can see the problem a day early; operators who catch and ignore them find out at the end.

Account Level Action

Login fails, the session is killed, or the number cannot re-register. This is not a restriction at all, it is a termination, and the appeal route is different. If the account still logs in, you are not in this category.

Terminations usually follow something categorical rather than a pattern: a number recycled from a previously terminated account, a report volume far outside the normal range, or content that crossed a line rather than a threshold. Appeals exist here too, but they are answered on different grounds, and the odds are poorer.

The Test That Tells You Which

Try to message one person who already has you in their contacts. If that works and a stranger does not, you have a spam restriction or a cooldown. If nothing works but the account opens, look at the wait value you are getting, which our error guide explains in full. If the account does not open at all, stop reading this section and treat it as terminated.

Do the test once, with one contact, and then stop. It is diagnostic, not exploratory. Running it repeatedly across a dozen strangers to see where the line sits is the behaviour that turns a reviewable restriction into an unreviewable one, and it produces no information you can act on.

Step One: Stop Everything

The first step is the one people skip, and skipping it is the single biggest reason restrictions become permanent.

Halt All Outbound Contact

Stop the campaign, stop the script, stop the manual sends. Not slow down, stop. Every additional attempt while restricted is another data point on the same pattern, and the system is still scoring while you are locked out. Operators who keep poking at it during the restriction period are the ones who never come back.

The same applies to the adjacent operations. Adding people to groups with Member Adder or anything like it, sending invite links in bulk, and joining a run of new groups all draw on the same reputation, and doing any of them while restricted reads as continuing rather than pausing. If it involves reaching people who have not reached you, it stops.

Take the Account Out of the Pool

If the account sits in a rotation, remove it rather than leaving it in a paused state where something might pick it up. Automation that retries on failure is particularly dangerous here, because it will hammer the restriction every cycle without a human noticing.

Check the scheduler as well as the tool. Cron jobs, warm-up scripts, and anything that logs in on a timer will keep the account active during the exact window in which quiet is the whole strategy. Restrictions have been extended by a forgotten warm-up job that nobody remembered was running.

Do Not Start a Replacement Yet

The instinct is to spin up a fresh account and continue. Resist it for now. A new account performing the identical sequence from the same network, at the same pace, to the same list, is the easiest pattern in the world to detect, and it usually gets restricted faster than the first one did. Fix the pattern before you buy the replacement.

There is also a practical reason to wait beyond pattern detection. Until you have worked out what caused the restriction, a replacement is an experiment with no hypothesis. You will spend money on an account, lose it the same way, and end the week with two restricted accounts and the same unanswered question.

Step Two: Read the Exact Message

The precise wording carries information, and most people read it once in frustration and never look at it again.

Message Text Versus Error Code

If you are in the app, screenshot the message. If you are on the API, log the error name and any wait value that came with it. These are different views of the same event and they carry different detail, so record both when you have both. Our guide on Telegram add member limits maps the full error set if what you are seeing is not one of the two above.

Keep these records per account rather than per campaign. When a second account is restricted weeks later, the only way to tell whether you have a systemic problem or two unrelated incidents is to compare the two sets side by side, and by then nobody remembers the first one accurately.

Reconstruct the Last Hour

Write down what the account did immediately before the restriction: how many messages, over how long, to how many strangers, from what network, with what text. This is not paperwork. It is the only way to work out what to change, and by tomorrow you will not remember accurately.

Include the things that feel irrelevant: which network the session was on, whether anything else was running from the same origin, whether the message had a link in it, whether the list came from a source you had used before. Restrictions are usually produced by a combination, and the factor you leave out because it seems unimportant is often the one that repeats.

Look for the Report

Restrictions rarely arrive from volume alone. Somewhere in that last hour, one or more recipients pressed report. If your message was identical for everyone, it is worth reading it as a stranger would, because that text is the input that generated the reports.

Read it as somebody who did not ask for it and does not know you. If the first line is about you rather than about them, if the link arrives before the reason, or if the message would be equally true sent to anyone at all, you have found the input that generated the reports. Fixing that is worth more than any pacing change.

Step Three: Appeal Through the Spam Bot

Telegram provides an official route for exactly this situation, and it is the only route that exists. There is no support email for it, no form, and no paid escalation.

How the Appeal Works

The appeal is submitted from the restricted account itself through Telegram's own anti-spam bot. It is reviewed and it can be declined. It is a review, not a switch, and no tool or service changes that. Anyone selling an unlock is selling either the appeal you can submit yourself or nothing at all.

Submit it from the restricted account itself. An appeal sent from a second account about a first one has no standing, because the reviewer has no way to connect them and no reason to trust the connection. This sounds obvious and is nevertheless one of the most common wasted attempts.

What to Write

Keep it short, specific, and free of theatre. State what the account is for, state that you understand the restriction, and state what you have changed. Do not claim you never messaged strangers if you did. Do not send the same text from ten accounts, because ten identical appeals read as exactly what they are.

Length works against you here. A short factual note is easier to review and harder to disbelieve than several paragraphs of explanation, and explanation tends to slide into the kind of detail that answers questions nobody asked. Say what the account is for, acknowledge the restriction, say what changed, and stop.

Submit Once

One appeal, then wait. Repeated submissions do not raise priority and they add activity to an account that should look quiet. If the answer is no, the answer is no; a second identical appeal will not produce a different one.

If you run several accounts and more than one is restricted, resist the urge to submit for all of them in the same hour with the same text. Handle them separately and write each one about its own account, because a batch of identical appeals arriving together describes an operation rather than a mistake.

Step Four: Wait Without Testing

This step is entirely passive and it is where most recoveries fail, because doing nothing feels like doing nothing.

Why Testing Resets the Clock

Every attempt to message a stranger while restricted is the restricted behaviour, performed again, recorded again. The account is under closer observation during this window than at any other time. Curiosity is the most expensive thing you can spend here.

This is also the period in which people quietly try a second device, a different network, or a fresh client, reasoning that the restriction might be local. It is not, and each of those attempts adds a new origin to an account already under review. The account looks less like a person with a problem and more like something being probed.

What Ordinary Use Looks Like

If the account is going to be used again, it should behave like an account during this period: read things, reply to people who write in, exist. What it should not do is sit completely dark for days and then produce a burst of cold messages the moment the restriction lifts, which is the pattern that produces an immediate second restriction.

If the account has a real purpose beyond outreach, this is a good week to use it for that purpose. If it has no purpose beyond outreach, that fact is itself worth noticing, because an account whose entire history is one repetitive action is the profile the anti-spam system is built to find.

How Long

There is no published duration and anyone quoting one is describing a single account. Some clear in days. Some do not clear. Plan on the assumption that this account is out of service for a while, and if the campaign cannot survive that, the campaign was too dependent on one account to begin with.

Plan around the account being gone rather than around it coming back. If a single restriction stalls the whole operation, the operation was one account deep, and that is the finding worth acting on this week regardless of what the appeal returns.

Step Five: Come Back Differently

If the restriction lifts, the account is not restored to where it was. It is on notice, and the behaviour that got it restricted will restrict it again faster.

Start Far Below the Old Pace

Whatever volume the account was doing when it was restricted, come back at a fraction of it and climb across days rather than within a session. The first week after a lifted restriction decides whether the account has a second life or a second restriction.

A workable shape is to spend the first days on replies and existing contacts only, then introduce a small number of new conversations per day, then climb from there across a week or more. The exact numbers matter less than the direction: every day should be slightly busier than the last rather than a return to the previous normal.

Change the Input, Not Just the Speed

Slowing down a message that generated reports produces slower reports. If the text was generic, the list was cold, or the offer made no sense to the recipient, those are the things that need fixing. Pace buys room around the report rate; it does not lower it.

The clearest signal you have improved is a change in what comes back. Fewer blocks, more replies that engage with the substance, and questions instead of silence all indicate the message is landing as something a person might want. Report rate is the number that restricted the account, and it is the number that has to move.

Assume It Is Watched

Treat a recovered account as one that has a mark against it, because it does. Use it for warm contacts, replies, and lower risk work, and put new cold volume on accounts that are not carrying history. This is a rotation decision rather than a settings decision.

Give the recovered account a defined role and write it down, otherwise it drifts back into the rotation the first time capacity is short. An account that quietly returns to cold volume three weeks later, because nobody remembered why it was set aside, is a restriction waiting to be rediscovered.

What Causes Restrictions in the First Place

Five inputs produce almost every restriction. None of them is volume on its own, which is why volume-only advice keeps failing.

Reports From Recipients

This is the dominant signal and everything else is secondary to it. A message that people report will restrict an account at low volume; a message people ignore politely will not, at much higher volume. If you change one thing, change what the message says and who receives it.

This is also the only factor that scales in the wrong direction as you grow. Doubling volume with the same message roughly doubles the reports, so every improvement in targeting or copy buys more room than any improvement in pacing. Operators who treat outreach as a delivery problem keep hitting the same wall from the same side.

Account History

An account with real history absorbs mistakes that a new one does not. History means exchanged messages, groups joined at human intervals, a filled profile, and activity spread over time. An account created this morning has no buffer at all, and its first burst is also its first impression.

History cannot be fabricated quickly, which is exactly why it works as a signal. An account that joins twenty groups in an hour and fills its profile in the same minute is describing itself as prepared rather than lived in, and preparation is the thing being detected.

Speed and Shape

Fifty messages across a working day and fifty messages in ten minutes are different events to a pattern detector, even though the total matches. Perfectly even intervals are their own signal, because people are not metronomes.

Time of day matters too. An account sending steadily through the night in its own registered timezone reads differently from one that works ordinary hours, and an account that never rests at all reads as a process rather than a person. Uneven, human, occasionally idle is the shape to aim for.

Identical Text

The same characters delivered to hundreds of strangers is trivially detectable and it is also what makes recipients reach for report. Variation helps on both counts, but only if it is real variation rather than a synonym swapped in the same sentence frame.

Real variation means different sentences, different openings, and different lengths, not the same skeleton with three interchangeable adjectives. The second kind is easy to generate and easy to detect, and it does nothing about the underlying problem that the message reads as bulk to the person receiving it.

Network Origin

Datacentre address ranges carry more suspicion than residential ones, and an account that hops between countries between sessions looks compromised rather than busy. The proxy guide covers the whole of this in detail.

Consistency matters as much as type. One account pinned to one origin, kept there, looks like a person with a home connection; an account that appears from a different country each session looks compromised, and the security response to that is not the same as the anti-spam response but it ends the campaign just as effectively.

Prevention: Warm-Up and Pacing

Everything above is recovery. Prevention is cheaper by a wide margin, and it is mostly a matter of patience in week one.

The First Week

Give a new account several days before it does any cold work. Photo, name, bio. Join a few groups that fit the account's supposed interests, spaced across days. Exchange real messages with real people. The target state is an account that would look unremarkable to anyone who opened it.

Resist automating the warm-up. A script that performs the ideal warm-up sequence produces exactly the regularity that gives it away, and the whole point of the week is to look unremarkable. Doing it by hand for a few minutes a day across several days is slower and works better.

Ramp Instead of Launch

Start well below what you believe the ceiling is and raise the volume across days, not within a session. Nothing about a slow first week is wasted, because the alternative is buying the account twice and losing the campaign in between. Our guide to avoiding bans in cold outreach goes through the full warm-up sequence.

Set the ceiling before you start rather than discovering it during the run. An operator who decides in advance to stop at a given number for the day stops there; an operator who decides to keep going while it works keeps going until it does not, which is by definition the point at which something has already gone wrong.

Watch the Early Signals

Small rate limits are the system telling you the pace is wrong while there is still time to fix it. Rising report rates, replies asking who you are, and blocks are all earlier and cheaper signals than a restriction. Most operators only look at the last one.

Build the check into the routine rather than the crisis. A weekly look at reply rate, block rate, and the smallest rate limits per account catches drift while it is still cheap, and it takes minutes. Nobody does this until they have lost an account, and everybody does it afterwards.

Fix the List Before the Pace

A well targeted list at moderate speed outperforms a cold list at any speed, because it produces fewer reports per message sent. This is a sourcing problem rather than a tooling problem; ProspectPulse exists on the discovery side for the same reason, and how you build the list is where it is solved.

Relevance is measurable in a way that intuition is not. Split a small run between two segments, keep everything else identical, and compare replies and blocks. Two hundred messages will tell you more about your targeting than another month of pacing adjustments.

When an Account Is Not Worth Saving

Recovery is not always the right call, and the sunk cost of a restricted account makes this decision harder than it should be.

The Honest Test

Ask what the account is actually worth: its age, its contacts, its group memberships, its number. A three year old account with a real number and a full contact list is worth waiting weeks for. A session created last Tuesday on a recycled virtual number is worth nothing, and time spent appealing for it is time not spent on the thing that produced the restriction.

Put a number on it before you decide. What did the account cost, what would replacing it cost, and how many hours are you about to spend on the appeal and the wait? For a cheap account that arithmetic answers itself immediately, and the time is better spent on the message that caused the problem.

Repeat Restrictions

An account restricted twice will be restricted a third time faster. At that point it is telling you it is finished for cold work, though it may still be perfectly useful for warm contacts and replies. Retire it from the front line rather than from the pool.

Keep a short record per account of what happened and when. Without it, a pool of a dozen accounts becomes impossible to reason about after a few months, and the account that has been restricted twice gets handed the next cold campaign because it happened to be idle.

Retire, Do Not Burn

Retiring an account does not mean deleting it. Its contacts, groups, and history stay useful, and a quiet account costs nothing to keep. What it should not do is carry cold volume again.

Retired accounts also have a second use as the warm side of your list. Existing conversations, group memberships, and contacts are exactly the relationships that make messaging safe, so the account that can no longer open doors can still walk through the ones already open. Our guide on Telegram engagement covers what to do with those relationships.

Capacity Comes From the Pool

Everything in this guide is about one account. The strategic answer is to stop depending on one.

The Arithmetic

Pushing a single account to ten times its safe volume means operating far outside safe pacing and eventually losing it, along with whatever was in progress. Spreading the same total across ten accounts keeps every one inside its comfortable range, and a problem with one costs a tenth of your capacity instead of all of it.

There is a resilience argument as well as a volume one. Ten accounts each doing a modest amount can absorb losing one without the campaign noticing, and they can be rested individually when their own signals turn. One account cannot be rested at all, because resting it stops everything.

Why It Collapses by Hand

Several accounts each need their own session, pacing, cooldown tracking, and network origin, and a mistake in one bleeds into the others through pattern similarity. The thing being managed stops being the campaign and becomes account health, which is a different job with different tools.

The failure mode is rarely dramatic. It is one account that quietly ran twice because two sessions were open, one that never rested because its cooldown was tracked in somebody's head, and one that shared an origin with three others for a week. Each is small and forgivable; together they produce a pattern that looks deliberate.

Where Floqal Fits

Account Manager holds the sessions, keeps each account inside its own observed limits, staggers work across the pool, and shows which accounts are healthy and which are resting. Mass DMs runs the sending against that pool with pacing and message variation built in rather than bolted on. Neither removes Telegram's limits, and any product claiming otherwise is describing something it cannot deliver. What they do is spend the limits well, which is the whole of the advantage and enough.

The measurable difference is in wasted work rather than raw speed. Requests that cannot succeed are not sent, flood waits are honoured rather than fought, and no single account carries a load that belongs to five. None of that is exciting and all of it is the difference between a pool that lasts months and one that lasts weeks.

Starting Without a Pool

If you are starting from nothing, OLD accounts remove the warm-up week and give you something with history from day one. What matters is not the age on the label but what the account has done, and whether the number behind it is real rather than recycled. Current availability is on the pricing page.

Whatever the source, verify before you build on it. Log in, look at what the account has actually done, and run a small warm test before it carries anything real. An account that fails quietly on its first cold batch will take the campaign with it, and the test costs an afternoon.

Frequently Asked Questions

How long does a Telegram spam restriction last?

Telegram does not publish a duration and it is not fixed. Some accounts clear in days, some sit restricted much longer, and some never clear. The only lever that exists is the appeal through Telegram's own anti-spam bot, which is a review rather than a switch and can be declined. Any specific figure you find quoted online is one person's account on one occasion, not a rule.

What does "you can only send messages to mutual contacts" mean?

It means the account has been restricted by the anti-spam system and can no longer initiate contact with people who do not have it saved as a contact. Existing conversations, groups, and replies still work normally, which is why the restriction is often missed until a campaign fails. At API level the same state appears as PEER_FLOOD.

Can I remove a Telegram restriction with a tool or a paid service?

No. The only route is the appeal from the restricted account through Telegram's own anti-spam bot, and it is reviewed by Telegram. Anyone selling an unlock is either submitting the appeal you could submit yourself or selling nothing. No proxy, client, or reseller changes the outcome.

Should I just make a new account instead of appealing?

Only if the restricted one is worth little: a session created recently on a recycled virtual number is not worth waiting for. But a new account repeating the identical sequence, from the same network, at the same pace, to the same list, usually gets restricted faster than the first one did. Fix the pattern before buying the replacement, whichever way you go.

Does using a proxy prevent spam restrictions?

It removes one signal, not the cause. Datacentre address ranges attract more suspicion than residential ones and hopping between countries between sessions looks like a compromised account, so a clean consistent origin helps. The dominant signal is still recipient reports, which are produced by the message and the list rather than by the connection.

Will the restriction come back after it lifts?

It will if the behaviour does. A recovered account is on notice and the same pattern restricts it faster the second time. Come back at a fraction of the old volume, climb across days rather than within a session, and change what the message says and who receives it, not only how quickly it goes out.