A versatile speech-to-text utility that leverages OpenAI Whisper to transcribe audio and video files into multiple subtitle and text formats.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install whisper-transcribe
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install whisper-transcribe using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Whisper Transcribe?
Whisper Transcribe is a high-performance transcription skill designed to convert audio and video content into structured text with exceptional accuracy. Built as part of the Openclaw Skills ecosystem, it utilizes OpenAI's Whisper models to provide features like automatic language detection and word-level timestamps. It is an essential tool for developers and content creators who need to process voice recordings, podcasts, or meeting notes into searchable, accessible formats.
This skill offers a flexible interface for different hardware environments, allowing users to choose between various model sizes depending on their RAM and precision requirements. By integrating this into your workflow, you gain the ability to handle a wide array of audio extensions including mp3, wav, m4a, and flac while generating standard outputs like SRT and VTT for immediate video integration.
Whisper Transcribe Use Cases
- Automatically generating subtitles for video content to improve accessibility and reach.
- Transcribing long-form podcasts or interviews into text for easier editing and SEO optimization.
- Converting recorded lectures and business meetings into organized notes and summaries.
- Batch processing large libraries of legacy audio files into structured JSON for data analysis.
- Translating and transcribing foreign language audio using the built-in auto-detection capabilities.
How Whisper Transcribe Works
- The user invokes the transcription script, passing one or more audio or video files as input.
- The skill identifies the requested model size, ranging from tiny for speed to large for maximum accuracy.
- FFmpeg is utilized in the background to decode the media files into a format readable by the Whisper engine.
- The system automatically detects the spoken language if a specific language code is not provided.
- The Whisper model processes the audio stream, identifying speech segments and converting them to text.
- The script formats the results into the requested output types (txt, srt, vtt, or json) and saves them to the specified directory.
Whisper Transcribe Setup
To get started with Whisper Transcribe from Openclaw Skills, you must install the core dependencies on your system:
# Install the OpenAI Whisper CLI
pip install openai-whisper
# Ensure ffmpeg is installed for audio decoding
# For macOS:
brew install ffmpeg
# For Ubuntu/Debian:
sudo apt update && sudo apt install ffmpeg
After installation, you can run the provided transcription script. Note that the first run will automatically download the required model files (e.g., approximately 150MB for the base model).
Whisper Transcribe Data Schema & Taxonomy
Whisper Transcribe manages its output by generating files that share the base name of the source audio. The data can be organized as follows:
| Format | Description |
|---|---|
| .txt | A clean, plain-text transcript of the spoken content. |
| .srt / .vtt | Standard subtitle files with precise start and end timestamps. |
| .json | A detailed data structure containing segments, word-level timing, and model metadata. |
| --output-dir | A configurable directory path where all processed transcripts are centralized. |
Whisper Transcribe Advanced Features
- Support for five different model tiers to balance processing speed and transcription accuracy.
- Word-level timestamping for granular alignment of text and audio.
- Batch processing support using wildcards to handle hundreds of files in a single session.
- Comprehensive output options allowing the simultaneous generation of all supported file formats.
- Seamless integration into larger automation pipelines within the Openclaw Skills framework.
SKILL.md
Loading
Related Openclaw Skills
Universal Notify
A unified notification script that sends alerts across multiple platforms like Telegram, ntfy, and SMTP through a single interface.
Pi Health Monitor
A lightweight diagnostics script to monitor thermal performance, power stability, and resource usage on Raspberry Pi hardware.
Dokploy Deployment Management
A comprehensive interface to programmatically manage Dokploy projects, deployments, and infrastructure through its native API.
Shelv
Shelv converts complex PDF documents into structured, searchable Markdown filesystems that can be navigated using standard Unix tools.
Outlook
A comprehensive toolset for managing Microsoft Outlook and Hotmail email and calendar services through the Microsoft Graph API.
x402-direct
A specialized directory for discovering and searching crypto-native, x402-enabled API services designed for autonomous agents.
Featured*
