whisper-cli

Transcribe local audio and video files into timestamped JSON transcripts using whisper.cpp.

3|1|Updated May 19, 2026

Search Tags:#offline #transcription #whisper #audio #speech-to-text #ggml

Ready to run

npx skills add https://github.com/adbertram/cli-tools --skill whisper-cli

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: whisper-cli
Source: https://github.com/adbertram/cli-tools/tree/main/_repo/skills/whisper-cli
Command: npx skills add https://github.com/adbertram/cli-tools --skill whisper-cli

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill removes the need for cloud-based transcription services by providing a secure, private, and offline interface for converting audio and video files into text.

Core Features & Use Cases

Local Transcription: Transcribe media files entirely on your local machine using ggml models without sending data to external servers.
Flexible Output: Generate JSON transcripts with timestamped segments or word-level precision, compatible with standard workflows.
Use Case: Quickly transcribe a local meeting recording or interview audio into a structured JSON format for documentation or analysis without internet connectivity.

Quick Start

Use the whisper-cli skill to transcribe the audio file located at path/to/audio.wav and save the output.

Frequently Asked Questions about whisper-cli

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

FAQPage Schema

How do I transcribe audio files offline on my local machine?▼

To transcribe audio files offline, you process media entirely on your local machine using a local ggml model and the whisper-cli binary. This removes the need for cloud-based transcription services and generates text without internet connectivity.

Does offline speech-to-text transcription work with video files?▼

Yes, offline speech-to-text transcription supports both audio and video file processing. You provide the media file path, and the whisper.cpp engine converts the spoken content into text without sending your data to external servers.

What output formats can I generate for timestamped audio transcripts?▼

Timestamped audio transcripts are generated as JSON files. You can produce transcripts with timestamped segments or extract word-level precision data, which makes the output compatible with standard documentation and analysis workflows.

Do I need a specific model to run local speech-to-text transcription?▼

Yes, local speech-to-text transcription requires a local ggml model to execute. You must configure this model alongside the whisper-cli binary to perform the transcription tasks without any network dependencies.

What is the best way to ensure private transcription of meeting recordings?▼

The best way to ensure private transcription of meeting recordings is using a local, offline interface. By converting your audio files into text entirely on your local machine, you guarantee that no data is sent to external servers.

Why use offline transcription instead of cloud-based speech-to-text services?▼

Offline transcription provides a secure and private environment for converting media into text. By using the local whisper.cpp engine with a ggml model, you avoid uploading sensitive audio or video files to cloud-based transcription services.