Back to Cookbook
OpenClaw recipe

Telegram Voice Transcription

Takes the confusion out of Speech-to-Text — already bundled, no install needed

Use Telegram voice memos to talk to your KiloClaw. The audio file is sent to the OpenAI Whisper API and your Claw reads the text and responds.

CommunitySubmitted by frankisokPersonal5 min
Try in KiloClawFree 7-day trial

INTEGRATIONS NEEDED

PROMPT

Set up the open-whisper-api skill with my API key: {YOUR_KEY_HERE}. Then add a rule to AGENTS.md: whenever I receive an audio file message from Telegram, automatically transcribe it using the open-whisper-api skill and respond to the transcribed text.

How It Works

Use Telegram voice memos to talk to your KiloClaw. The audio file is sent to the OpenAI Whisper API and your claw reads the text and responds.

What Others Get

There are many options available for enabling Speech-to-Text, so if you are overwhelmed with the choices this takes the confusion out of it. The skill is already bundled with OpenClaw so doesn't require an install. The API cost is just $0.006/minute of audio and it works with the Kilo free models.

Setup Steps

  1. Set this api key {YOUR_KEY_HERE}, for the open-whisper-api skill.
  2. Add a rule to AGENTS.md to remind you that any time you receive an audio file message from Telegram you are to use this skill to transcribe the audio.
Tags:#productivity#automation#integration#workflow