Package Exports

media-transcriber
media-transcriber/dist/index.js

This package does not declare an exports field, so the exports above have been automatically detected and optimized by JSPM instead. If any package subpath is missing, it is recommended to post an issue to the original package (media-transcriber) to support the "exports" field. If that is not possible, create a JSPM override to customize the exports field for this package.

Readme

media-transcriber

Batch transcribe audio/video files using pluggable AI backends (Whisper local, OpenAI API, and more).

Quick Start

# Run directly with npx (no install)
npx media-transcriber setup

# Or install globally
npm install -g media-transcriber

# Transcribe files (stateless)
media-transcriber transcribe ./data/input ./data/output

Requirements

Node.js >= 18
FFmpeg and ffprobe in PATH
- Windows: winget install ffmpeg or scoop install ffmpeg
- macOS: brew install ffmpeg
- Linux: sudo apt install ffmpeg

Backend requirements:

whisper-local: Python 3.10+ with openai-whisper installed
whisper-api: OpenAI API key (OPENAI_API_KEY env var or --openai-api-key)

Usage

Transcribe

# Basic
media-transcriber transcribe ./recordings ./transcripts

# Model and device
media-transcriber transcribe ./data/input ./data/output -m medium -d cpu

# OpenAI API backend
media-transcriber transcribe ./data/input ./data/output -b whisper-api --openai-api-key <key>

# Keep temporary files in ./data/output/temp
media-transcriber transcribe ./data/input ./data/output --include-temp

# JSON output for agents
media-transcriber transcribe ./data/input ./data/output --json

Setup

Dependency check wizard (no persistent config file):

media-transcriber setup

Execution Options

All parameters are passed at execution time (stateless CLI).

Field	Type	Default	Description
`inputFolder`	string	required	Input folder argument
`outputFolder`	string	required	Output folder argument
`backend`	string	`whisper-local`	Transcription backend
`whisperModel`	string	`large-v2`	Model name
`device`	`cuda` or `cpu`	`cuda`	Processing device
`maxDurationSeconds`	number	`1200`	Split threshold
`enableAudioEnhancement`	boolean	`false`	Enable enhancement filters
`keepIntermediateFiles`	boolean	`false`	Keep temp files with `--include-temp`
`tempFolder`	string	`<outputFolder>/temp`	Temp working folder
`pythonPath`	string	unset	Python path for local backend
`openaiApiKey`	string	env/flag	API key for OpenAI backend

AI Agent Integration

Structured JSON

media-transcriber transcribe ./data/input ./data/output --json 2>/dev/null

Progress Events (stderr)

In --json mode, NDJSON progress events are emitted to stderr.

Exit Codes

Code	Meaning
`0`	Success (all files transcribed)
`1`	General error
`2`	Missing dependency
`3`	Configuration/argument error
`4`	No input files found
`10`	Partial success

Supported Formats

Input: .m4a, .mp3, .mp4, .mkv, .wav, .flac, .ogg, .webm

Output: .txt, .srt

Development

npm install
npm run typecheck
npm test
npm run build

License

MIT