parakeet

Transcribe audio files to timestamped text using parakeet-mlx on Apple Silicon.

Updated Dec 9, 2025
One-click install
npx skills add https://github.com/SecKatie/kmtools --skill parakeet
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: parakeet
Source: https://github.com/SecKatie/kmtools/tree/main/parakeet
Command: npx skills add https://github.com/SecKatie/kmtools --skill parakeet

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Transcribing audio to text on-device eliminates reliance on cloud services and speeds up access to searchable transcripts.

Core Features & Use Cases

  • On-device transcription with Parakeet-MLX for fast, private speech-to-text
  • Timestamped transcripts suitable for podcasts, interviews, and meetings
  • Works with common formats like mp3, wav, and m4a; outputs text and SRT

Quick Start

Transcribe an audio file by running uvx parakeet-mlx /path/to/audio/file.mp3

Frequently Asked Questions about parakeet

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I transcribe a podcast to text on-device?

To transcribe a podcast to text on-device, use Parakeet-MLX to process audio files locally. It accepts formats like mp3 and outputs timestamped text without relying on cloud services.

Does on-device audio-to-text work with Apple Silicon?

On-device audio-to-text requires Apple Silicon hardware to run Parakeet-MLX. You also need the uvx tooling and an initial 2.5GB model download to generate transcripts.

Can I get timestamped transcripts for meetings from mp3 files?

You can get timestamped transcripts for meetings from mp3 files by running the transcription tool. It processes common audio formats and outputs text or SRT files with timestamps.

What audio formats are supported for local speech-to-text?

Local speech-to-text supports common audio formats including mp3, wav, and m4a. The tool processes these files on-device to deliver searchable text transcripts.

Why does on-device transcription require a 2.5GB download?

On-device transcription requires a 2.5GB download to fetch the Parakeet model, enabling local processing. This eliminates cloud reliance and ensures private, fast audio-to-text conversion.