Most comparison pages about Telegram bots are vague about money. Group Help isn't: the free tier is free forever, Pro is 150 Stars once and covers five groups, UltraPro is 150 Stars a month per group. You can decide in thirty seconds whether it fits your budget, which is more than most tools let you do.
What that money buys is worth listing honestly, because the answer to "should I switch" depends on whether you are using any of it.
Before you pick a Group Help alternative: what 150 Stars gets you
The free tier already covers moderation commands, welcome, goodbye and rules messages, anti-spam, anti-flood, banned words, custom commands, one recurring message, and a clone bot you can rename and give your own avatar. Pro adds the anti-porn and NSFW filter, a rule requiring members to have a profile picture, auto-deletion filtered by message type and role, a block on accounts that join and leave quickly, and automatic replies. UltraPro is where it gets unusual — force-join requirements, keyword auto-replies, silent punishments, a second banned-words list with a different penalty, banned words checked inside usernames, twenty custom roles with custom permissions, fifty recurring messages.
Twenty custom roles is not a feature you build unless real communities asked for it. GroupHelp is a genuinely deep moderation product, and nothing below is an argument that it isn't.
Captcha: Group Help, Telegram, and what it actually stops
Most Group Help bot Telegram setups turn this switch on first, and fairly so. Its join protection and captcha do the job they claim: an account running a join script never answers the check, so it never reaches the chat. If your attackers are cheap bought bots, that single switch removes most of them.
The accounts that cost people money are the other kind. Aged, real profile photo, plausible username, banned from eleven other channels last month. It clears a captcha without breaking a sweat, sits quietly for a week, and then does what it came for. There is nothing inside your group that could have known any of that.
A captcha asks "are you human". A network list asks "which eleven channels banned you last month". Only one of those questions can be answered from inside a single group.
Three filters at the door instead of one
TGuard checks an arriving account three separate ways. There is a captcha, same as Group Help's. There is a join-rate limit you configure — more than ten joins in thirty seconds and the entire burst gets kicked or banned, which is the only thing that works against a raid, because a raid is forty accounts arriving before anybody types a word. And there is the abuser list: fifteen million accounts, built from bans across 15,000+ protected channels, with the flag raised once an account has been banned in five or more of them.
On the message side TGuard has caught up with the things Group Help has always had. Six rules per chat, each configured on its own settings page: stop words with substring or whole-word matching, a link filter for http://, bare domains and t.me / telegra.ph, a frequency rule defaulting to five messages in ten seconds, separate media and audio rules, and the known-abuser rule. Each carries its own action — record, delete, or delete and ban — and each starts in record-only mode, because the fastest way to lose a healthy chat is a stop word that fires on ordinary conversation.
Then there is the half nobody's group bot covers
Group Help protects groups. That is the name, and it is accurate. But most Telegram operations in 2026 are a channel with a comment thread attached, and the channel is where the money is: subscriber count sets your ad rate, ERR sets whether anyone pays it, and a wave of bought subscribers wrecks the second while inflating the first.
Nothing that reads messages can see that. There are no messages in a channel — only posts you wrote and an audience you did not vet. TGuard's channel side handles the subscriber audit, bulk cleanup of bots already inside by join date, fake view detection, scam label and shadow ban monitoring, and giveaway and ad-campaign protection.
| Job | Group Help | TGuard |
|---|---|---|
| Banned words, links, anti-flood | Yes, free tier | Yes |
| Captcha at entry | Yes | Yes |
| Night mode, custom roles, recurring messages | Yes — the core of the paid tiers | No |
| Cross-network ban history | No | 15M+ accounts, 15,000+ channels |
| Join-rate limit on a burst | Quick join-and-leave block (Pro) | Whole burst, threshold you set |
| Channel subscribers, ERR, fake views, scam label | No | Yes |
What TGuard doesn't do
No night mode. No custom roles, no recurring messages, no anti-porn classifier, no welcome or goodbye texts, no clone bot with your own name and avatar. If your group runs on any of that, keep Group Help — the honest recommendation for a large community is both bots, with delete rights given to one of them only. Two bots racing to delete the same message produce confusing logs and the occasional survivor, because each assumed the other had taken it.
Who should switch
- A group whose problem is text and the behaviour of real people: Group Help, free tier. There is nothing to add here.
- A group being flooded with accounts in bursts: TGuard. The captcha is not the point — the join-rate limit acting on the whole burst is, and so is the network ban history. Neither can be assembled from inside one group.
- A channel with subscribers you sell ads against: TGuard, starting with an audit rather than with settings. Group Help does not reach here by design, not by quality.
- A channel plus its comment thread: TGuard in both the channel and the linked discussion group, rules on record-only for the first day.
Setting it up
- Add @channel_guardian_bot to the group as an admin with delete and ban rights — and to the linked discussion group as well if you run a channel with comments.
- Turn on Spam Protection, then set a join-rate limit that matches your normal growth. This is the setting that pays for itself the first time you are raided.
- Leave every message rule on record-only for a day, read what it caught, then promote the ones you trust to delete or ban.
- If bots are already inside, run a scan and clear them in bulk rather than one at a time.
For the wider picture, the companion piece on choosing a Telegram anti-spam bot covers what the free options cover and where they run out, while the Combot comparison looks at the same split from the analytics side and the ChatKeeper one from the depth-of-configuration side.
Frequently Asked Questions
For spam and bot defence, yes. TGuard runs stop words, a link filter, a frequency rule, media and audio rules, plus captcha, join-rate limits and a shared abuser list. For moderation comfort — night mode, custom roles, recurring messages, anti-porn filtering, welcome and goodbye texts — no. Group Help keeps those, and running both bots is a normal setup.
A captcha stops a join script. It does not stop an aged account with a real profile photo that clears the check and posts something bland for a week first. TGuard uses three entry filters instead of one: the captcha, a join-rate limit that acts on the whole burst, and a network list that flags any account banned across five or more of 15,000+ protected channels.
Yes, if you add it to the linked discussion group — that is where Telegram keeps comments, and it is a separate chat from the channel. What no group bot covers is the channel itself: subscriber quality, fake views, boosts and the scam label are not messages, so nothing that reads messages can see them.