Every article about Telegram automation eventually quotes a number: so many messages an hour, so many joins a day. Almost none of those numbers come from Telegram. Knowing which is which saves accounts.
What Telegram does state
A handful of structural limits are documented and stable: how many channels one account can create, how many members a group holds, file size caps. These are product facts, not anti-spam thresholds, and they are the same for everyone.
What is never published
Anti-spam thresholds. How many messages to strangers trigger a restriction, how fast is too fast, how many reports it takes. Publishing those would hand spammers a specification, so they are deliberately secret and deliberately variable.
They also depend on the account: age, history, whether it has a photo and a name, whether people reply to it. Two accounts doing the same thing on the same day get different outcomes, and both experiences end up online as "the limit".
How folklore forms
Someone runs twenty messages an hour for a week without trouble and writes it down as a rule. Someone else repeats it with a fresh account on a shared proxy and loses it in a day. Neither observation is dishonest; both are unrepeatable.
Working without numbers
- Grow gradually. Whatever the ceiling is, walking towards it beats starting at it — warming up accounts.
- Vary the pace. Even intervals are a stronger signal than volume — quiet hours.
- Watch the account, not the counter. The reliable signal is the account's own state — checking for a spam block.
- Spread the work. Volume comes from more accounts, not harder ones — spreading load.
The one number that matters
Not a threshold but a ratio: how many accounts you lose per month against how much they produce. That figure is yours, it is measurable, and unlike anything you read online it describes your setup rather than someone else's.
How Neurogram does this
Everything above is manual work that runs into the number of hours in a day. Neurogram does it on your own accounts and without you: it collects channels and audiences, writes comments and replies with a language model, and keeps limits and quiet hours so the accounts survive. See pricing or look at your case.