All nodes/Integrations/Telegram

Telegram Group History

Reads the recorded history of a Telegram chat: the messages the bot observed in a group or channel, including those that never triggered a workflow. Needs both the 'record group history' switch in the deployment AND privacy mode disabled on the bot (or admin rights) — otherwise the bot simply never sees the messages. A blank Chat ID means the chat of the running turn. Outputs: Messages (for LLMs), Records (raw rows carrying message_id/file_id) and Transcript (plain text). This is the group's history, not the current message: the message that started this run — with its voice note, photo or document — arrives on the Entry node's Message output; wire that straight into Speech → Text or the model.

Try it

When to use it

Reach for this node when the scenario needs to know what actually happened in a group — including the messages that never started a turn. This is not the bot’s own conversation: Chat History returns the assistant’s dialogue with one person, while this is the many-voiced feed of a group or channel. When you need a keyword rather than “the last N messages”, use Search Messages. The message that started the scenario does not need this node: it arrives — with its voice note, photo or document — on the Message output of Entry.

How it works

The Bot API cannot hand back past messages — there is no “show me yesterday’s conversation” method. So this node reads the platform’s own recording and never calls Telegram at all. Recording is switched on in the deploy panel, under “Group history → Record group history”, and it is off by default. The second condition is the bot’s privacy mode: while it is on, Telegram shows the bot only commands, replies and mentions, so there is almost nothing to record. See the Telegram channel page.

The node has three outputs. Messages is an array an LLM node accepts as is: every line is prefixed with its author, bots arrive as the assistant role and everyone else as user. Records are the raw rows carrying message_id, media_file_id, date and author. Transcript is the same text flattened into one string.

Common mistakes

• An empty result is not an error. The run still succeeds. There are usually three causes: recording is off, privacy mode is on, or the window and filters excluded everything. Start with the diagnostics button in the deploy panel.
• The limit counts backwards. With Since 7d and a limit of 50 you get the last 50 messages of that week, not the first. The output itself is oldest first.
• Chat ID here is numeric only. @channelusername is not accepted (unlike the send node). A blank field means the chat of the running turn, so a run started from web chat, a schedule or the API stops with an error.
• Messages carries no file_id. To download an attachment, take media_file_id from the Records port and pass it to Get File.
• Storage is finite. It is a ring buffer (“Messages kept per chat”) plus your plan’s retention period — old messages disappear quietly.
• Deletions are only half-visible. Messages removed by the Moderate node drop out of the results. Deletions made by a human inside Telegram do not: the Bot API never reports them.

Inputs

—

Outputs

Configuration