name: groq-whisper description: Fast cloud speech-to-text via Groq Whisper API (replaces local openai-whisper). homepage: https://console.groq.com/docs/speech-text metadata: {"clawdbot":{"emoji":"🎙️","requires":{"env":["GROQ_API_KEY"]}}}

Groq Whisper — Audio Transcription

Name: groq-whisper
Rating: 88
Author: laurenzseifried

Transcribe audio files using the Groq API (whisper-large-v3-turbo).

Quick Start

bash

node skills/groq-whisper/transcribe.js /path/to/audio.mp3
node skills/groq-whisper/transcribe.js /path/to/audio.m4a --language de
node skills/groq-whisper/transcribe.js /path/to/audio.mp3 --translate --format txt

Programmatic

const { transcribe } = require("./skills/groq-whisper/transcribe");
const result = await transcribe("/path/to/file.mp3", { language: "de" });
console.log(result.text);

Models

Model	Use Case
whisper-large-v3-turbo	Default transcription (fast)
whisper-large-v3	Translation / higher accuracy

Env

•GROQ_API_KEY — Required. Get from https://console.groq.com/keys

Supported Formats

mp3, mp4, mpeg, mpga, m4a, wav, webm — max 25 MB.