Paid vs free Telegram moderation bots: cost, limits and exit

Stay on a free moderation plan while it covers your essential rules, your moderators can manage it and its data practices are acceptable. Pay when a specific limit or missing workflow creates enough work or risk to justify the cost. Group size alone does not settle that decision.

A subscription buys access to the provider's stated service. It does not buy a clear community policy, a guarantee against false positives or a human appeal process. Budget for those separately.

What “free” can mean#

A permanently free public bot, a free tier and a trial of paid features are different offers. A trial may make a workflow look affordable until it expires. A free tier may cover your moderation needs but impose an unrelated constraint, such as advertising or shorter reporting history.

Official pages reviewed on October 5, 2026 provide these starting points. We did not create accounts, purchase plans or verify account-specific checkout totals.

Public offer sources: Rose FAQ, Combot pricing and ChatKeeper plan guide. Defendy's implementation was reviewed separately; the subscription guide is available for navigation.

Combot labels Pro for non-commercial use and Business for commercial use, and lists AI billing separately. Its advertised monthly amounts are not a verified total including any account-specific charges. ChatKeeper's English site presents dollar figures while its Russian plan guide presents ruble figures. Do not silently convert them into an equivalent offer or assume which one applies to you.

ChatKeeper's plan guide also describes a seven-day introductory period followed by Free, with advertising in greetings on the free plan. That is a concrete reason to review the post-trial state before the group becomes dependent on its initial configuration.

Price the work that remains#

Write down the fee and the time needed for setup, routine changes, false-positive review, reporting and handover. Add any separate hosting or engineering work if you are evaluating a self-managed alternative. Keep one-time setup effort separate from recurring work.

Here is a fictional planning example, not a price or performance claim about any product:

• Free setup: no service fee and three hours of moderator work each month
• Paid setup: a $15 monthly fee and one hour of work each month
• At a planning value of $20 per hour, the recurring totals would be $60 and $35

The paid option only has that advantage if the time saving actually happens. If both take three hours, its total becomes $75. For volunteers, the assigned hourly value is a way to discuss scarce time; it is not necessarily a cash expense.

Do not fill the time column with a vendor's promise. Observe your team's real workload or mark the estimate as uncertain. A service can also justify a fee through a necessary export or a particular control, even if it does not save hours. Name that requirement instead of disguising it as efficiency.

Which limit are you paying to remove?#

Build a short list of essentials and attach each to a documented plan entitlement. Useful questions include:

• Is the offer for one group, several groups or an account?
• Does the needed rule, report or integration belong to this tier?
• Is a quota measured per month, per group, per message or per operation?
• Does a limit stop processing, reduce available history or add a charge?
• Can a second administrator use the workflow without a separate purchase?

“No limit” on one resource does not remove platform limits or guarantee availability. Likewise, a longer history window does not establish that activity from before installation can be reconstructed. For reporting-specific decisions, use the analytics selection guide.

For Defendy, the reviewed subscription page loads available plans and the selected group's subscription, and includes a Telegram Stars purchase workflow. This verifies implementation, not a live invoice, current amount, renewal policy or production availability. Read the subscription guide, then rely on the actual offer available to your group before committing.

Read the end of the subscription before the beginning#

The exit questions are often more important than a launch discount. Establish whether renewal is automatic, how to stop it, what happens to existing rules and when access to historical data changes. Save the terms and plan description that apply to your purchase.

Combot's published refund policy describes cancellation through Billing and generally non-refundable payments, with stated exceptions for duplicate payments and a mistaken purchase for another group within 48 hours. Check the current policy yourself before prepaying a long period. Do not assume another service follows the same terms.

If downgrade behavior is undocumented, ask the provider for the exact effect on your essential rules. There is no need to purchase or deliberately exceed a quota just to discover a limit. Until the answer is clear, treat continuity after the paid period as unresolved.

Make a small exit pack#

Keep the community's policy, allowed exceptions, role assignments and configuration notes in a place the team controls. Where supported, export a sample and open it outside the service. Identify which data is included and which is missing.

Rose's export documentation describes settings export and import within its workflow. This does not establish import compatibility with other bots. More generally, an analytics CSV, a settings file and an incident journal solve different recovery problems.

Limit who can open exports containing member-level data. Before removing a service, check its retention policy and any separate deletion procedure. Also arrange a human moderation fallback; do not leave the group unprotected while waiting for migration answers.

The purchase decision in one sentence#

Pay for a named capability or a demonstrated reduction in work, after confirming the applicable offer and exit conditions. Otherwise, keep the free setup and improve the rules or handover first. Revisit the decision when your workload or requirements change, rather than upgrading because an article calls a tier “professional”.