Define how your connected Telegram bot answers messages with a visual flow.
When a user messages your connected Telegram bot, the active Telegram flow can read the message, use AI or conditions, ask for another user reply, and send a final response.
The bot connection and an active Telegram Flow are both required. Without an active Flow, incoming messages are ignored and the bot does not reply. Responses are controlled by the Flow; the integration does not answer directly from project settings.
Connect your bot first: Telegram Bot Integration. Then create and activate a flow with the Telegram trigger. Each project can have one active Telegram flow.
To verify the connection, build Telegram Input → Telegram Response, enter a short test message in Response Text, save the Flow, and activate it. Send a text message to the bot and check the Flow execution logs. Then add AI, knowledge-base search, or other processing as needed.
Telegram flows can use shared nodes such as RAG Search, Invoke LLM, conditions, filters, variables, text operations, and debug nodes.
Add Read Memory and Append Memory if the Flow needs conversation history. Incoming messages are processed in the background; the resulting reply is sent to the same Telegram chat. This connector accepts text messages; attachments receive an unsupported-content notice.
Existing bot connections now use Flows too. Keep the connection and activate a Telegram Flow in the same project. The former direct-answer mode is no longer available. If a Flow fails, review its execution logs; it does not switch to a direct chat answer.
Test the flow in a private chat before enabling it for a wider audience.