n8n Telegram Workflow: Transcribe Voice with Whisper-1

Transcribe Telegram voice messages to text with n8n + Whisper-1

This n8n workflow listens for new messages in Telegram, automatically distinguishes between text and voice, and converts voice notes into a clean transcript using OpenAI’s whisper-1—then sends the resulting text onward.

What this workflow does

• Triggers on new Telegram messages: The workflow starts with the Telegram “Message Trigger” node whenever a message is received.
• Routes content by type: A Switch node checks the incoming message:
  • If the message is text, it follows the text branch.
  • If the message contains voice, it follows the voice branch.
• Downloads Telegram audio: For voice messages, the workflow retrieves the audio file from Telegram.
• Transcribes using OpenAI: The audio is sent to an OpenAI Transcribe Recording node that uses the whisper-1 speech-to-text model to produce a transcript.
• Sends the final text: Whether the input was original text or a transcribed voice message, the final text is passed to a Send transcription message node.

Use cases

• Voice-to-text capture: Turn Telegram voice notes into text for logging or review.
• AI follow-ups: Pipe transcribed content into an AI agent (e.g., a large language model) for analysis or response generation.
• Operations & support: Convert spoken updates from team or customers into readable notes automatically.

Technical details

• Telegram integration: Telegram “Message Trigger” receives incoming messages.
• Routing logic: Switch node differentiates text vs voice.
• Speech-to-text: OpenAI Transcribe Recording using whisper-1.
• Send output: Send transcription message node forwards the transcript/text.
• Required setup: Telegram Bot Token (configured in the Message Trigger node) and an OpenAI API Key.