n8n Telegram Voice Transcription with Palatine Speech

Turn Telegram voice notes into instant text replies—automatically

This n8n workflow listens to your Telegram bot for incoming messages, transcribes any voice note using Palatine Speech (palatine_large_highspeed), and sends the transcription back in the same chat. Regular text messages are echoed back—so your bot always responds.

What this workflow does

• Telegram Trigger: Starts when your Telegram bot receives a new message.
• Message type check: Uses a Switch step to determine whether the message contains a voice note or text.
• Voice note transcription: If the message is a voice note, the workflow downloads the audio from Telegram.
• Palatine Speech transcription: Sends the audio to Palatine Speech using the model palatine_large_highspeed, then waits for the transcription result.
• Reply in Telegram: Posts the transcription text back to the same Telegram chat.
• Text echo: If it’s a text message, it replies by echoing the same text back.

Use cases

• Instant voice-to-text support in Telegram, where users speak and receive readable transcripts right away.
• Hands-free updates for teams or SaaS ops—record voice notes and instantly capture the text.
• Quick testing and prototyping for n8n Telegram bots that need voice transcription.
• Operational logging setups (optional customization) by storing every transcript in Google Sheets or a database.

Technical details

• Telegram bot integration: Requires Telegram Bot API credentials (BotFather token) and a publicly reachable webhook.
• Palatine Speech integration: Uses the community node n8n-nodes-palatine-speech with Palatine Speech API credentials.
• n8n nodes used: Telegram Trigger, Switch, Telegram, plus a Sticky Note for guidance, and the Palatine Speech node.

Simple setup: Configure Telegram + Palatine Speech credentials, activate the workflow, then send a voice note to confirm it replies with the transcription.