ChatKeeperBot — the ChatKeeper bot Telegram admins usually mean when they say the name — publishes a tier table, and it is more honest than most landing pages: 12 triggers on free against 240 on ADVANCED, an action log capped at 1,000 entries against unlimited, an ad in the greeting that goes away once you pay. You can see exactly what the money buys — depth of chat configuration.
Which is also why comparing it to TGuard head-on misses the point. We ask a member different questions.
Before you pick a ChatKeeper alternative: what it does well
Thirty-two filters: words, media, voice messages, emoji, regular expressions, 18+ content, face control on username and display name. Triggers with 25 condition types and 16 action types on free, rising to 45 and 30 on ULTIMATE. A reputation system with ranks and experience, a referral programme, forms and surveys, recurring messages, eleven statistics graphs and an engagement map of members. It runs in MAX as well as Telegram.
This is a product about the life of a chat — who is active, who deserves a rank, which word your filter catches most. If you run a community where people talk to each other rather than just read posts, half that list will matter to you, and none of it is duplicated in TGuard.
Two questions that never overlap
A filter, a trigger and a reputation score all answer "what is this person writing, and how useful are they". That is the right question for exactly as long as there is a person on the other side.
When four hundred accounts join a channel overnight, the question is different: "who are these, and how did they get in". No filter answers it, because there is nothing to read — those accounts never wrote anything. Padding the counter was the whole job.
A reputation score starts counting from a member's first message. Half the trouble in Telegram comes from accounts whose first message never arrives.
Both have a captcha — count it in threes
ChatKeeper's verification types are richer: math captcha, fixed captcha, subscribe to a channel, a form with validation. One caveat — two of them are available on the free tier, and all four unlock on ADVANCED.
TGuard has one verification type, and on its own it is not better. The difference is that the captcha does not stand alone: next to it sits a join-rate limit you configure yourself — more than ten joins in thirty seconds and the whole burst gets kicked or banned — and a network list where an account is flagged after being banned in five or more of 15,000+ protected channels. Fifteen million accounts are in that database today.
An aged account with a real profile photo clears any captcha, validated form included. It does not clear the third filter, and that filter cannot be built from inside a single chat at all. It needs the network.
What TGuard runs on the message side
Six rules per chat, each with its own action — record, delete, or delete and ban. Stop words with substring or whole-word matching: whole-word keeps "cast" from firing on "podcast", substring catches the deliberately mangled spellings. A link filter for http://, bare domains and t.me / telegra.ph, off by default because plenty of healthy chats are built on people sharing links. A frequency rule defaulting to five messages in ten seconds. Separate media and audio rules. And the known-abuser rule, which needs no keyword at all.
Every new rule starts in record-only mode. You watch what it would have deleted for a day before you let it delete anything, because the fastest way to kill a healthy chat is one careless stop word firing on ordinary conversation. Each rule also has a cap on how often it can fire per window: a rule that goes berserk pauses itself and writes a warning instead of clearing the room.
This is the same class of job as ChatKeeper's 32 filters, and on depth of configuration we lose. We win in one specific place — the abuser rule also runs in the comments under channel posts, which a bot only reaches through the linked discussion group. In one ordinary 28-minute window on our engine, 561 of 1,749 messages from already-banned accounts landed exactly there.
And the channel side, which a chat bot does not have
| Job | ChatKeeper | TGuard |
|---|---|---|
| Message filters | 32 types | 6 rules |
| Triggers, reputation, forms, referrals | Yes — the core product | No |
| Verification types at entry | 4 types, 2 on free | 1 type + join-rate limit + network list |
| Cross-network ban history | No | 15M+ accounts, 15,000+ channels |
| Cleanup of bots already inside | No | In bulk, by join date |
| Channel subscribers, ERR, views, scam label | No | Yes |
| MAX messenger | Yes | No |
The last four rows are the border. A channel has no message stream: it has subscribers, reach, an ad rate and a scam label it can pick up. A wave of bought subscribers wrecks ERR while inflating the counter, and the next advertiser closes the deal cheaper. No filter and no trigger sees that, because there is nothing there to read. Hence the subscriber audit, fake view detection and anti-raid.
What TGuard doesn't do — a longer list than usual
No triggers. No reputation with ranks and experience. No referral programme, no forms or surveys, no recurring messages, no engagement map, no eleven graphs of member activity. No MAX. If your chat lives on any of that, TGuard will not replace it and does not try — run both bots and give message-delete rights to one of them only.
If you arrived looking for a ChatKeeperBot alternative, the fork below is more honest than any “top 10” list.
Picking one
- A living community where people talk to each other: ChatKeeper, and the free tier covers a lot if you can live with the ad in the greeting.
- A channel with subscribers you sell ads against: TGuard, and start with an audit of how many bots are already inside rather than with settings.
- A channel plus its comment thread: TGuard in both the channel and the linked discussion group, rules on record-only for the first day.
- A large community with both problems: both bots, delete rights to TGuard, and clear the accumulated bots in bulk rather than one at a time.
The comparisons with the other two neighbours in this niche — Combot and Group Help — cover the same split from the analytics and configuration sides, and the anti-spam roundup covers what free bots catch and where they stop.
Frequently Asked Questions
For spam and bot defence, yes: stop words, a link filter, a frequency rule, media and audio rules, captcha, join-rate limits and a shared abuser list. For the life of a chat, no. TGuard has no triggers, no reputation system with ranks and experience, no forms, no referral programme and no engagement map. That is deliberate — TGuard is about who arrived, not about how they behave once they are in.
ChatKeeper offers richer verification types — math captcha, fixed captcha, subscribe-to-channel, a validated form — but only two of them are on the free tier; all four unlock on ADVANCED. TGuard has one verification type, with two more filters standing next to it that a single-chat bot cannot have: a join-rate limit acting on the whole burst, and a network list flagging any account banned across five or more of 15,000+ protected channels.
Yes, and for a large community that is the normal setup: ChatKeeper runs the chat — triggers, reputation, statistics — while TGuard covers the door and the channel side. One rule: give message-delete rights to one bot only. Two bots racing to delete the same message produce confusing logs and the occasional survivor.