Audio transcription tool. Converts audio files to text using Whisper, with optional LLM post-processing as needed. Use when: Use when: transcribing meeting recordings, podcast transcripts, or when text extraction from audio files is required.
Installs into .claude/skills of the current project.
Are you the author of Transcribe Tool?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/xuiltul-transcribe-tool)
---
name: transcribe-tool
description: >-
Audio transcription tool. Converts audio files to text using Whisper, with optional LLM post-processing as needed.
Use when: Use when: transcribing meeting recordings, podcast transcripts, or when text extraction from audio files is required.
tags: [audio, transcription, whisper, external]
---
# Transcribe Tool
Audio transcription tool using Whisper (faster-whisper).
## Invocation
**Bash**: Run with `animaworks-tool transcribe transcribe <音声ファイル> [オプション]`
### audio — Audio Transcription
```bash
animaworks-tool transcribe transcribe audio_file.wav [-l ja] [-m large-v3-turbo]
```
## Parameters
| Parameter | Type | Default | Description |
|-----------|------|---------|-------------|
| audio_path | string | (required) | Path to the audio file |
| language | string | null | Language code (e.g., ja, en). Auto-detected if null |
| model | string | "large-v3-turbo" | Whisper model name |
| raw | boolean | false | If true, skips LLM post-processing |
## CLI Usage
```bash
animaworks-tool transcribe transcribe audio_file.wav [-l ja] [-m large-v3-turbo]
```
## Notes
- faster-whisper must be installed
- When using GPU, CUDA-compatible ctranslate2 is required
- The model is automatically downloaded on first run
