Answers as the message arrives
Telegram delivers each message to your bot’s webhook, so a reply goes out when the model has written it, not on a schedule.
Describe your Telegram bot in plain words and AI builds it. It answers each message the moment it arrives, replies with a model, and can look up an order, add a lead or post to a channel through any of 1,400+ tools. No message chains to draw, and nothing to host.
The same three steps for a support assistant, a lead catcher or a channel poster.
In plain words: what it answers, which of your apps it may use, and when to hand over to a person.
AI writes the bot once, as code. The builder shows it as a map of its steps before anything runs.
Publish the bot, then point your Telegram bot at the address Whenever gives you. Every message after that is answered as it arrives.
Real briefs for the same kind of job. Pick one and Whenever writes it against your accounts; the code is yours to change from there.
How a Telegram bot on Whenever answers, and what it costs.
Telegram delivers each message to your bot’s webhook, so a reply goes out when the model has written it, not on a schedule.
A run is one message handled, however many tools the bot calls to answer it. Model tokens draw on the AI allowance every plan includes, then are billed on top.
The bot does not remember earlier conversations between runs. Give it the facts it needs through its tools: an order lookup, a sheet, a CRM record.
Yes. Create the bot with BotFather, describe what it should do to Whenever in plain words, and Whenever writes it, connects the token and runs it. You never have to write the code.
Yes. It replies with a model and chooses which of the tools it was given to use, such as an order lookup or a lead form. It cannot reach a tool its code did not offer it.
No. Whenever hosts it and gives you the webhook address to register with Telegram once. There is nothing of yours to keep online.
AI building is free within weekly usage limits. Once it is live, each message it handles is one run, and model tokens use the separate runtime AI allowance your plan includes before they are billed on top.
Sources reviewed September 30, 2026.
Describe what it should answer and what it may do. Whenever writes it and runs it.