openai-whisper

Transcribe audio files to text locally using the Whisper CLI.

Updated Feb 16, 2026

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

Ready to run

npx skills add https://github.com/travis-burmaster/agentbox --skill openai-whisper-travis-burmaster

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: openai-whisper
Source: https://github.com/travis-burmaster/agentbox/tree/main/agentfork/skills/openai-whisper
Command: npx skills add https://github.com/travis-burmaster/agentbox --skill openai-whisper-travis-burmaster

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Transcribing audio recordings typically requires uploading files to cloud APIs, which raises privacy concerns and recurring costs. This Skill runs OpenAI's Whisper model locally via its CLI, converting speech to text without an API key or internet dependency. ## Core Features & Use Cases - Local Speech-to-Text: Transcribe audio files like MP3 and M4A entirely on your machine with no API key required. - Translation to English: Use the `--task translate` option to translate non-English audio into English subtitles. - Flexible Output Formats: Generate plain text or SRT subtitle files, with selectable model sizes to trade speed for accuracy. - Use Case: A podcaster records an interview as an M4A file and needs a transcript for show notes. Run Whisper locally with the medium model to produce a text file without sending the recording to any external service. ## Quick Start Transcribe the attached audio file 'interview.mp3' to a text file using Whisper with the medium model.

Frequently Asked Questions about openai-whisper

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

FAQPage Schema

How do I transcribe audio to text with Whisper CLI?▼

Run `whisper /path/audio.mp3 --model medium --output_format txt --output_dir .` to transcribe an audio file locally. The output text file is written to the specified directory, and no API key is needed.

How to translate audio to English subtitles using Whisper?▼

Use the `--task translate` flag, for example `whisper /path/audio.m4a --task translate --output_format srt`. Whisper translates non-English speech into English and writes the result as an SRT subtitle file.

Does Whisper transcription require an internet connection or API key?▼

No API key is required because Whisper runs entirely locally. Internet access is only needed once to download the model files, which are cached in ~/.cache/whisper for subsequent runs.

Which Whisper model size should I use for transcription?▼

Smaller models run faster but with lower accuracy, while larger models are slower and more accurate. This install defaults to the turbo model; use `--model medium` or larger when accuracy matters more than speed.

Why is Whisper transcription slow on the first run?▼

The first run downloads the selected model weights to ~/.cache/whisper, which can take time depending on model size. Later runs reuse the cached model and start transcribing immediately.