This workflow creates a smart Telegram chatbot that can handle both text and voice messages, interact with an AI language model, and reply to users seamlessly. Designed for enhanced conversational engagement, it transcribes voice messages into text for processing, uses GPT-4 to generate responses, and then delivers those responses back to the user in Telegram. The workflow uses nodes for message listening, content type determination, voice transcription via OpenAI, and message management, providing a fluid multi-format chat experience.

The process begins with listening for any incoming messages on Telegram. It then checks if the message contains text or voice. If a voice message is received, the audio file is downloaded and transcribed into text using OpenAI’s speech recognition. Both text and transcribed voice messages are then combined and sent to an AI model (GPT-4) for generating contextual replies. The response, with error handling for unsupported commands, is then sent back to the user, completing the interactive loop.

This workflow is particularly useful for creating a multi-format support chatbot that can handle voice messages effectively, providing a more natural and versatile communication interface for users, especially in customer support or personal assistant scenarios.

Node Count	11 – 20 Nodes
Nodes Used	@n8n/n8n-nodes-langchain.agent, @n8n/n8n-nodes-langchain.lmChatOpenAi, @n8n/n8n-nodes-langchain.memoryBufferWindow, @n8n/n8n-nodes-langchain.openAi, set, stickyNote, switch, telegram, telegramTrigger