Converts audio and video files into accurate text transcripts with precise word-level timestamps using WhisperX.
The Audio Transcriber skill enables Claude to perform high-quality speech-to-text conversion directly within your development environment. By leveraging WhisperX, it provides multi-language support and superior word-level alignment compared to standard Whisper implementations. Whether you need to generate SRT subtitles for a video, transcribe a recorded meeting, or extract structured JSON data from speech, this skill automates the entire process including audio extraction from video files and VAD filtering to ensure clean results.
主要功能
01Support for various formats including MP3, WAV, MP4, and MKV
02High-precision word-level timestamp alignment
03Multiple model sizes ranging from 'tiny' for speed to 'large-v2' for accuracy
04Multi-language support with automatic speech detection
05Export options for TXT, SRT, VTT, and structured JSON
060 GitHub stars
使用场景
01Transcribing technical meetings or interviews into searchable documentation
02Extracting dialogue from media files for programmatic data analysis
03Generating professional subtitle files (SRT/VTT) for video content
