Real-time AI Translation

Watch Any Video in Your Language

VoiceToSub automatically detects spoken language and translates it into English subtitles. Download any foreign video, pick the file from the menu bar, and get it back with embedded subtitles — completely offline using on-device AI.

Get Started FreeSee How It Works
youtube.com
🎬
"Welcome to our show, today we will discuss..."
Japanese → English

Powerful Features

Everything you need for seamless video translation

🌍

99 Languages Supported

Auto-detects Japanese, Chinese, Korean, Spanish, French, German, and 93 more languages.

⚡

Real-Time Translation

Subtitles appear as you watch with minimal delay. Choose between local processing or fast cloud API.

🔒

Privacy First

Run 100% locally with no data leaving your machine, or use OpenAI's API — your choice.

🎬

Translate Video Files

Pick any video file from your Mac, translate it offline, and get an MKV with embedded optional English subtitles.

🎯

Works on Any Video

Translate movies, TV shows, YouTube downloads, or any local video file. No existing captions needed — subtitles are generated from the audio.

🔧

Choose Your Model

Five Whisper models from tiny (75 MB, fastest) to large-v3 (3 GB, most accurate). Switch anytime from the menu bar.

💰

No Subscriptions

Free to use locally. For cloud API, you pay OpenAI directly at cost — no markup, no subscription.

🛠️

Open Source

Fully transparent code. Self-host the server, customize it, contribute to it.

How It Works

Get started in minutes

1

Download the Desktop App

Download the VoiceToSub macOS app — a lightweight menu-bar application. On first launch it guides you through setup and downloads the Whisper AI model (~500 MB).

2

Pick a Video to Translate

Click "Translate Video" in the menu bar, pick any video file, and choose local AI (free, offline) or OpenAI cloud (faster). Progress shows in real time.

3

Get Subtitled Output

VoiceToSub produces an MKV file with embedded English subtitles, revealed in Finder when done. Play it in VLC, IINA, QuickTime, or any video player.

Three Ways to Use

Choose what works best for you

🍎

Desktop App

  • ✓Translate local video files
  • ✓MKV output with optional subtitles
  • ✓Choose from 5 Whisper models
  • ✓100% offline, no browser needed
  • ✓Built-in onboarding tutorial
🖥️

Local Processing

  • ✓100% free, no API costs
  • ✓Complete privacy — nothing leaves your Mac
  • ✓Works offline
  • −Requires desktop app (free download)
  • −Slower on CPU (~15-30s delay)
☁️

OpenAI Cloud

  • ✓Fast — only 5-8 second delay
  • ✓No GPU needed
  • ✓Best translation quality
  • −Pay-per-use via OpenAI — no markup
  • −Requires internet + API key

Get Started

Free and open source. Up and running in minutes.

1

Download the Desktop App

Native macOS menu-bar app — free, no account needed

VoiceToSub is a native macOS menu-bar application. Pick any local video file, choose your AI model, and get an MKV with embedded optional English subtitles — no internet required in local mode.

🍎 Download for macOS

Works on Intel and Apple Silicon (via Rosetta 2)

How to install

  1. Open the downloaded .dmg file
  2. Drag VoiceToSub.app into your Applications folder
  3. Launch VoiceToSub from Applications — a quick tutorial walks you through the features
  4. A ◌ VoiceToSub icon appears in your menu bar while the model loads (first run downloads ~500 MB)
  5. Once it shows ● VoiceToSub the server is ready

Desktop app features

  • Translate Video — pick any local video file and get a subtitled MKV
  • Model Selection — switch between tiny, base, small, medium, and large-v3
  • Source Language — set the source language for better accuracy
  • Privacy — local mode processes everything on your Mac, no data leaves your machine

macOS says it can't be opened?

Open System Settings → Privacy & Security and click "Open Anyway". This is a one-time step for unsigned apps.

2

Start Translating

Click the menu-bar icon → Translate Video → pick a file. Progress shows in the menu bar. When done, the subtitled MKV opens in Finder — ready to play in VLC, IINA, or any video player.

Support the Project

VoiceToSub is free and open source. If it helps you, consider buying me a coffee to support development.

Buy Me a Coffee