Ready-made recipe · ai

Telegram → AI Agent → Telegram

An AI Agent answers each Telegram message, looking things up in Google Sheets when it needs to, and replies on Telegram.

Steps

When when a telegram message comes in — Telegram node only reads
Then answer it using google sheets as a tool — AI Agent node · may use: Google Sheets can be undone
Then reply on telegram — Telegram node can’t be undone

Fields to fill (5)

The file has every node wired up. These are the blanks only you can fill (your form, your sheet, your words).

Node	What to do
Telegram	fill `chatId` (Unique identifier for the target chat or username, To fin...)
Telegram	fill `text` (Text of the message to be sent) — the AI step's output: run that node once in n8n and pick its output field
OpenAI Chat Model	check model: OpenAI Chat Model chosen by default (most used) — swap for Gemini, Anthropic, Ollama … if you like
OpenAI Chat Model	fill `model` (The model)
Google Sheets tool	pick `workflowId`: a small workflow that does the Google Sheets part (Execute Workflow Trigger → the Google Sheets action(s) the agent may take); the node list we read (n8n 2.43.2) has no Google Sheets tool node to plug in directly

Accounts to connect (4)

In n8n, pick or add a credential on each node — one of the types below.

Telegram: telegramApi
Telegram: telegramApi
OpenAI Chat Model: openAiApi
Google Sheets tool: googleApi or googleSheetsOAuth2Api (in that small workflow)

Need something slightly different? Same engine, your own words:

Run it on Apify — type your own sentence, get the workflow back in seconds. Open Bridgecore on Apify (pay per use, $0.02–$0.05; nothing charged on errors).
Ask your AI — add the MCP server https://mcp.usebridgecore.com/ to Claude, Cursor or any MCP client, then just say what should happen. Tools: do, find_recipe, plan_automation, make_n8n_workflow. Free to plan.
Ask me first — add “and ask me first” to any step (“… and post it on X and ask me first”) and the workflow sends you an Approve / Decline message (Slack, Telegram or Gmail) before that step runs. Nothing is sent, posted, deleted or paid until you tap Approve; no answer in a day = declined.

Get the n8n file

Download telegram-agent-answers-from-sheets.json

Open a new, empty workflow in n8n.
Open the workflow menu (⋯) and choose Import from File — or copy the JSON below and paste it onto the canvas (Ctrl/Cmd+V).
Connect the accounts, fill the fields above, run one test, then turn the workflow on.

Or copy the JSON

{
  "name": "An AI Agent answers each Telegram message",
  "nodes": [
    {
      "id": "n1",
      "name": "Telegram Trigger",
      "type": "n8n-nodes-base.telegramTrigger",
      "typeVersion": 1.5,
      "parameters": {
        "updates": [
          "message"
        ]
      },
      "position": [
        240,
        300
      ],
      "notes": "step: when a telegram message comes in · reversibility: read-only"
    },
    {
      "id": "n2",
      "name": "AI Agent",
      "type": "@n8n/n8n-nodes-langchain.agent",
      "typeVersion": 3.1,
      "parameters": {
        "promptType": "define",
        "text": "=Answer it:\n\n{{ JSON.stringify($json) }}"
      },
      "position": [
        460,
        300
      ],
      "notes": "step: answer it using google sheets as a tool · reversibility: reversible"
    },
    {
      "id": "n3",
      "name": "Telegram",
      "type": "n8n-nodes-base.telegram",
      "typeVersion": 1.2,
      "parameters": {
        "resource": "message",
        "operation": "sendMessage",
        "chatId": "",
        "text": ""
      },
      "position": [
        680,
        300
      ],
      "notes": "step: reply on telegram · reversibility: irreversible"
    },
    {
      "id": "n4",
      "name": "OpenAI Chat Model",
      "type": "@n8n/n8n-nodes-langchain.lmChatOpenAi",
      "typeVersion": 1.3,
      "parameters": {
        "model": {
          "__rl": true,
          "mode": "list",
          "value": ""
        }
      },
      "position": [
        460,
        500
      ]
    },
    {
      "id": "n5",
      "name": "Google Sheets tool",
      "type": "@n8n/n8n-nodes-langchain.toolWorkflow",
      "typeVersion": 2.2,
      "parameters": {
        "description": "Call this to use Google Sheets.",
        "workflowId": {
          "__rl": true,
          "mode": "list",
          "value": ""
        }
      },
      "position": [
        640,
        500
      ]
    }
  ],
  "connections": {
    "Telegram Trigger": {
      "main": [
        [
          {
            "node": "AI Agent",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "AI Agent": {
      "main": [
        [
          {
            "node": "Telegram",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "OpenAI Chat Model": {
      "ai_languageModel": [
        [
          {
            "node": "AI Agent",
            "type": "ai_languageModel",
            "index": 0
          }
        ]
      ]
    },
    "Google Sheets tool": {
      "ai_tool": [
        [
          {
            "node": "AI Agent",
            "type": "ai_tool",
            "index": 0
          }
        ]
      ]
    }
  },
  "settings": {
    "executionOrder": "v1"
  },
  "active": false
}