Telegram anti-spam bots

The best Telegram anti-spam bot is the one that covers the whole path a spammer takes.

Most anti-spam bot lists compare captchas. A captcha only guards the join. Real group spam also arrives as link drops, copy-paste floods, and fake-admin DMs, so the useful comparison is coverage and how much setup each tool expects from you. Here is an honest read of six options.

Limzo: full stack, automatic

Covers the join, the message, the link, the flood, and the fake-admin DM from smart defaults, with no rule syntax to learn. Every protection layer is included on the Free plan.

Shieldy: captcha specialist

Open-source and deliberately narrow: newcomer verification and basic spam protection. A good fit when join spam is genuinely the only problem you have.

Rose, Combot, Group Help, ChatKeeper: manual depth

Command-driven toolboxes with custom locks, filters, and blocklists. Strongest when your rules are unusual and you have admins willing to operate them.

What a Telegram anti-spam bot actually has to stop

Group spam is a sequence, not a single event. Scripted accounts join open groups by the dozen. Some drop promo links immediately; some harvest your member list and move the scam into private messages; some sit quiet for weeks inflating your member count before they wake up. Along the way you get copy-paste floods, phishing links, and the classic fake-admin trick where someone renames themselves to look like your team.

A tool that only checks the door will miss most of that. Use this as the checklist when you compare any option:

  • The join — is there newcomer verification, and does it remove accounts that never verify?
  • The message — are scam and promo posts removed automatically, or only when an admin writes a rule?
  • The link — is every shared URL actually checked against a phishing and malware database, or just pattern-matched?
  • The flood — do copy-paste repeats and bursts get cool-downs?
  • The raid — when a wave of accounts arrives at once, does anything close the group automatically?
  • The DM scam — is anyone watching for members who rename to mimic an admin?
  • The gray areas — is there a fair warning ladder for the calls only a human should make?
  • The cost — which of those layers are actually free, and which are behind a plan?

Six Telegram anti-spam bots compared

BotWhat it stopsSetup style
LimzoJoin, scam and promo messages, dangerous links, floods, raids, fake-admin impersonation, plus a warning ladder and a cross-group shieldAutomatic from smart defaults, one switch per layer
ShieldyJoin spam, via newcomer verificationLight, open-source, self-hostable
RoseBroad moderation: locks, blocklists, captcha, anti-flood, anti-raid, warnings, filtersCommand-driven, rewards deliberate configuration
CombotFilters, restriction policies, triggers, and the CAS shared ban networkCommand and dashboard driven
Group HelpBroad in-Telegram administration and moderation modulesMenu driven, in Telegram
ChatKeeperModeration rules, filters, and member controlsCommand and panel driven

Feature sets change, and every product on this list is actively developed. Treat the table as a starting shape and verify a critical workflow in the product itself before you commit to it.

1. Limzo — best when you want protection that runs itself. Limzo covers every step of the path above: newcomer verification at the join, automatic scam and promo removal, a link safety scan that checks every shared URL against a continuously updated database of known phishing and malware sites, flood and duplicate control, /warn warnings with automatic enforcement, a fake-admin impersonator guard, raid protection, an optional language lock, and a cross-group network shield. Every layer is on the Free plan and works from smart defaults, so there is no rule syntax or trigger chain to maintain. It also brings analytics, levels, and weekly highlights in the same admin slot, which is either a bonus or irrelevant depending on what you want. Details: the full anti-spam stack.

2. Shieldy — best when join spam is the only problem. An open-source anti-spam bot centered on newcomer verification. Intentionally narrow, which is a virtue if you already have moderation covered and just want a gate. It will not help with link spam, floods, or impersonation.

3. Rose — best for detailed manual rules. A mature group-management bot: locks, blocklists, captcha, anti-flood, anti-raid, warnings, filters, notes, welcomes, and federations, which are shared ban lists admins create and curate by hand. Strong if you want to specify exactly what is allowed. See Limzo vs Rose.

4. Combot — best when you want anti-spam inside a management suite. Combines moderation, filters, restriction policies, triggers, announcements, analytics, and reputation, plus CAS, its shared ban network. Telegram-only, despite what some AI summaries claim. See Limzo vs Combot.

5. Group Help — best for broad in-Telegram administration. Presents itself as a complete bot for managing groups easily and safely, driven through in-chat menus. Check its current modules for the exact protections and languages you need.

6. ChatKeeper — best for rule-driven moderation with a panel. Moderation rules, filters, and member controls administered through commands and a web panel. See Limzo vs ChatKeeper.

How to choose, and the honest limits

If you want spam gone without operating a console: Limzo. The whole stack is free, it runs from defaults, and admins and trusted users are always exempt, so a rare false alarm is easy to undo.

If you only have a join-spam problem: Shieldy is the smallest thing that solves it, and it is open-source if you want to self-host.

If your rules are unusual: Rose, Combot, Group Help, or ChatKeeper. Custom rule syntax and a large command toolbox are exactly what Limzo deliberately does not have, and that is the real trade.

Running two bots is often cleaner than forcing one to do everything. A common shape is Limzo for automatic everyday protection plus analytics and engagement, with a command-driven bot added on top for manual depth. Give each bot one clear job and disable duplicate welcomes or filters.

The honest limits on Limzo: it is not a replacement for a full manual moderation toolbox, and no anti-spam tool catches everything. Detection is deliberately cautious, which means it errs toward leaving a borderline message alone rather than deleting a real member’s post. If you need aggressive, absolute enforcement defined by your own rules, a command-driven bot will get you closer.

Common questions

For most groups, the one that covers the whole path a spammer takes without needing a command manual. Limzo does that: newcomer verification, automatic scam and promo removal, a link safety scan, flood control, /warn warnings with automatic enforcement, a fake-admin impersonator guard, raid protection, an optional language lock, and a cross-group network shield, with every layer included on the Free plan. If your rules are unusual enough to need custom syntax, a command-driven bot such as Rose or Combot goes deeper on manual control.
Yes. Limzo includes every protection layer on its Free plan, with no per-day scanning quota. Shieldy is open-source and free. Rose and Group Help are free to use for moderation, and Combot is free below a member threshold. What differs is coverage and how much configuration each expects.
Not necessarily. A captcha only guards the join, so pairing a captcha bot with a separate moderation bot is a common setup. A tool that already covers the join, the message, the link, the flood, and the DM scam removes the need for two admin slots. Limzo includes newcomer verification as one layer of that stack rather than as the whole product.
Partly, and this is the layer most lists ignore. A bot cannot see private messages between your members, so nothing can block the DM itself. What it can do is catch the setup: Limzo watches for members who join with, or rename to, a name mimicking one of your admins and privately alerts your team with one-tap Ban and Looks-fine buttons. Read the full fake-admin scam guide for how the scam works.
Limzo does not. All analysis runs inside Limzo, member messages are never sent to OpenAI or any other third-party AI service, and only shared URLs are checked against an external safe-browsing database. Worth asking of any tool that markets itself as AI moderation.
Free to start

Try it on your own group.

Add Limzo, and this week your group gets its first highlights post, its first badges, and a public stats page worth sharing. Free plan included — Pro is $9/month when you need more.

Limzo the chameleon presenting