Files
AKharytonchyk 8137cd03a9 Initial commit: ComfyUI Telegram Bot custom nodes
- Add TelegramListener node for receiving messages
- Add SaveToTelegram node for sending responses
- Include comprehensive documentation and examples
- Set up proper project structure with requirements
2025-06-23 00:23:19 +02:00

1.8 KiB

Example Workflow: Telegram Echo Bot

This example demonstrates how to create a simple echo bot using the Telegram custom nodes.

Workflow Steps:

  1. Telegram Listener Node

    • Input your bot token
    • Set timeout to 30 seconds (or desired wait time)
    • This will output the received message text and chat ID
  2. Processing (Optional)

    • You can add any processing nodes here
    • For example: text transformation, AI processing, image generation, etc.
  3. Save to Telegram Node

    • Use the same bot token
    • Connect the chat_id from the listener
    • Connect your processed message (or just echo the original)

ComfyUI Workflow JSON

{
  "1": {
    "inputs": {
      "bot_token": "YOUR_BOT_TOKEN_HERE",
      "timeout": 30
    },
    "class_type": "TelegramListener",
    "_meta": {
      "title": "Telegram Listener"
    }
  },
  "2": {
    "inputs": {
      "text": [
        "1",
        0
      ]
    },
    "class_type": "CLIPTextEncode",
    "_meta": {
      "title": "Echo the message"
    }
  },
  "3": {
    "inputs": {
      "bot_token": "YOUR_BOT_TOKEN_HERE",
      "chat_id": [
        "1",
        1
      ],
      "message": [
        "1",
        0
      ]
    },
    "class_type": "SaveToTelegram",
    "_meta": {
      "title": "Send back to Telegram"
    }
  }
}

Setup Instructions:

  1. Replace YOUR_BOT_TOKEN_HERE with your actual bot token from BotFather
  2. Load this workflow in ComfyUI
  3. Queue the workflow
  4. Send a message to your bot in Telegram
  5. The bot will echo the message back to you

Advanced Usage:

You can extend this workflow by:

  • Adding image generation nodes between the listener and sender
  • Processing the text through AI models
  • Adding conditional logic based on message content
  • Storing conversation history
  • Adding multiple response types