video-to-text-local

Transcribe YouTube URLs and local media to text and SRT on Apple Silicon.

40|6|Updated Mar 26, 2026
One-click install
npx skills add https://github.com/0xmariowu/Autosearch --skill video-to-text-local
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: video-to-text-local
Source: https://github.com/0xmariowu/Autosearch/tree/main/autosearch/skills/tools/video-to-text-local
Command: npx skills add https://github.com/0xmariowu/Autosearch --skill video-to-text-local

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires mlx_whisper.

What problem does it solve?

Transcribe video or audio locally to text and subtitle data without sending data to cloud services, preserving privacy and reducing costs.

Core Features & Use Cases

  • On-device transcription and subtitle generation using yt-dlp for audio extraction and local mlx-whisper inference.
  • Accepts YouTube URLs and local media files (audio and video) and outputs raw text plus SRT-style subtitles.
  • Ideal for video notes, podcast transcripts, offline archives, and privacy-conscious workflows on Apple Silicon devices.

Quick Start

Call transcribe with a YouTube URL or local media path to begin on-device transcription immediately.

Frequently Asked Questions about video-to-text-local

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

FAQPage Schema
How do I transcribe a video offline on Apple Silicon?

To transcribe video offline on Apple Silicon, this Skill extracts audio using yt-dlp and runs local mlx-whisper inference to generate text and SRT subtitle files entirely on-device.

Can I generate SRT subtitles from a YouTube URL locally?

Yes, you can generate SRT subtitles from a YouTube URL locally by passing the URL to the transcription function, which fetches audio via yt-dlp and processes it on-device with mlx-whisper.

Does on-device transcription with mlx-whisper require an internet connection?

On-device transcription with mlx-whisper does not require an internet connection for processing local media files, but fetching audio from YouTube URLs requires network access for the initial yt-dlp download.

What is the best way to create podcast transcripts privately without cloud services?

The best way to create podcast transcripts privately is using on-device transcription, which processes local audio files via mlx-whisper to ensure data never leaves your machine or touches cloud services.

How do I select a different mlx-whisper model for local transcription?

You can select a different mlx-whisper model for local transcription by setting the appropriate environment variable before running the Skill, allowing you to balance accuracy and performance.

Why does local video transcription fail with missing dependency errors?

Local video transcription fails with missing dependency errors when yt-dlp or mlx_whisper are not installed in your environment, which this Skill handles with robust error messaging to guide installation.