🔬 Audio transcription with Telegram and Groq Whisper

⚡ 928 views · 🔬 Document Extraction & Analysis

Description

Short description

Transcribe Telegram voice/audio messages to text using Groq’s OpenAI-compatible Whisper endpoint. Replies are delivered either as a Telegram message or as a downloadable .txt file, plug-and-play for n8n with minimal setup.

Who’s it for / Uses

• Educators, podcasters, interviewers, and support teams who need quick voice → text conversions.
• Automating meeting notes, voice feedback, voicemail transcription, or speech logging.
• Useful when you want transcripts pushed to chat or saved as files for archiving.

How it works (overview)

1. Telegram Trigger — workflow starts on incoming message.
2. Switch (Audio/Voice) — detects voice or audio. If neither, replies “Wrong file type” and stops.
3. Telegram Download — downloads the audio using the file_id, outputs file path/binary.
4. Set Node (credentials + options) — stores Groq_API and Telegram_access_token (required) and transcript_output_format (message or file).
5. HTTP Request → Groq (Whisper) — uploads audio (multipart/form-data) to Groq’s transcription endpoint and receives text.
6. Reply Switch — routes to either:
   • Message branch: send transcribed text as a Telegram message.
   • File branch: convert transcript to .txt and send as a document.

Requirements

• n8n instance (cloud or self-hosted) with internet access
• Telegram bot token (create via BotFather)
• Groq API key (create at https://console.groq.com/keys)
• Basic n8n nodes: Telegram Trigger, Switch, Telegram Download, Set, HTTP Request, Convert to File, Telegram Send Message/Document

Important setup & notes

• Mandatory: Add Groq_API and Telegram_access_token in the Set node (or use n8n Credentials). The workflow will fail without them.
• Do not hardcode keys in HTTP node fields that will be exported/shared. Use Set fields or n8n Credentials.
• Include sticky notes explaining each node (yellow note with full description recommended). Sticky notes should show setup steps and required fields.
• Before publishing: remove personal IDs and secrets, test with sample voice messages, and verify Groq response schema to map the transcript field correctly.

Security & best practices

• Use n8n Credentials or environment variables in production.
• Rotate API keys if they become exposed.
• Keep the Set node private when sharing templates; instruct users to replace keys with their own.

🔗 Nodes Used

HTTP Request, Telegram, Telegram Trigger, Convert to File

📥 Import

Download workflow.json and import into n8n: Workflow menu → Import from File

📖 Importing guide · 🔑 Credential setup