🌐 macOS · Apple Silicon

Live audio translation. Entirely on your Mac.

Whisper transcribes. Mistral 7B translates. Both run locally on Apple Silicon — no cloud, no API keys, no data leaves your machine. Pick any audio source, choose a target language, and go.

Whisper (base) Mistral 7B 4-bit MLX Apple Neural Engine ScreenCaptureKit 101 Languages 100% Local

Download How It Works ↓

LLT menu bar app — audio source, languages, overlay controls

LLT floating translation overlay — original text and translation

Audio Sources

Mic on the table. Zoom call. Any app. All work.

Choose your audio source and LLT captures it. WebRTC, Teams, a mic on a conference table — it doesn't matter. If your Mac can hear it, LLT can translate it.

🎙️

Microphone CoreAudio

Any hardware mic, USB mic, or audio interface. Select your device from the dropdown. Put a mic on a table and translate a meeting room in real time.

📱

Single App Audio ScreenCaptureKit

Capture audio from one specific app — Zoom, Teams, Chrome, Discord, FaceTime, or any other app. Only that app's audio gets translated, nothing else.

🔊

System Audio ScreenCaptureKit

Capture all system audio at once. Everything playing on your Mac gets transcribed and translated. Useful as a catch-all for multi-source scenarios.

Processing Pipeline

Audio in, translation out. All local.

The backend starts with the app and runs a Python server on localhost. Audio is captured, chunked, sent via WebSocket, transcribed by Whisper, translated by Mistral, and displayed as an overlay — all without leaving your Mac.

Audio Capture

Mic, app, or system audio → 3-second chunks via WebSocket to local backend

Whisper STT

Speech-to-text with auto language detection. Resampled to 16kHz. Runs on Neural Engine.

Mistral Translation

Mistral 7B Instruct (4-bit MLX) translates transcribed text to target language

Overlay Display

Floating overlay shows original + translation. Auto-hides after 6s. Optional TTS.

Floating overlay — original text (gray) + translation (bold white)

Menu Bar App

Lives in your menu bar. Starts on demand.

LLT is a menu bar app — no dock icon, no window clutter. Left-click opens the control panel, right-click shows status and quit. The backend starts with the app (or on demand), but translation only begins when you press Start.

🟢

Backend Lifecycle

Backend auto-starts with the app (configurable) and auto-stops on quit. Loads Whisper + Mistral models on startup (~6 seconds). Status visible in the control panel and menu bar icon.

▶️

Translation On Demand

Backend running ≠ translating. You manually press Start when you need translation. This keeps resource usage minimal until you actually need it — no background processing when idle.

LLT Settings — General tab with backend and display options

LLT Settings — multilingual interface (DE, EN, ES, FR, IT)

Settings — backend control, display options, and 5 interface languages

Features

Everything you need. Nothing in the cloud.

Overlay

Floating overlay at the bottom of your screen. Shows original text (small) and translation (large). Auto-hides after 6 seconds. Rounded corners, dark semi-transparent background.

Show Original Text

Display the original transcribed text above the translation in the overlay. See exactly what Whisper heard and how Mistral translated it.

Save Transcript

Save the full session as a timestamped TXT file in ~/Documents/LiveTranslator/. Original + translation for every segment. Also supports SRT subtitle export.

Text-to-Speech

Read translations aloud using macOS system speech. Useful when you can't look at the screen — the translation plays through your speakers or headphones.

Auto-Gain

Automatically adjusts input gain to maintain consistent audio levels for Whisper. Targets -14 dB RMS with smooth adaptation. Manual gain also available (0.5x–4x).

Auto-Start Backend

Optionally start the backend automatically when the app launches. Otherwise, start it manually from the control panel. Backend always stops cleanly on quit (SIGTERM → SIGINT → SIGKILL).

Settings — Info tab with version details and backend status

Languages

101 languages. Auto-detection included.

Whisper detects the source language automatically. Mistral translates to any of 101 target languages. Set source to "Auto" and just let it figure out what's being spoken.

🇩🇪 🇬🇧 🇪🇸 🇫🇷 🇮🇹 🇵🇹 🇳🇱 🇵🇱 🇷🇺 🇺🇦 🇹🇷 🇬🇷 🇨🇳 🇯🇵 🇰🇷 🇸🇦 🇮🇳 🇮🇱 🇹🇭 🇻🇳 🇮🇩 🇲🇾 🇸🇪 🇩🇰 🇳🇴 🇫🇮 🇨🇿 🇸🇰 🇭🇺 🇷🇴 🇧🇬 🇭🇷 🇷🇸 🇸🇮 🇱🇹 🇱🇻 🇪🇪 🇮🇸 🇬🇪 🇦🇲 🇦🇿 🇰🇿 🇿🇦 🇰🇪 🇪🇹 🇳🇬 🇧🇩 🇵🇭 🇭🇰

LLT — 101 languages available, Whisper recognizes, Mistral translates

101 built-in languages — no installation required, auto-detection included

Afrikaans, Albanian, Amharic, Arabic, Armenian, Assamese, Azerbaijani, Bashkir, Basque, Belarusian, Bengali, Bosnian, Breton, Bulgarian, Burmese, Cantonese, Catalan, Chinese, Croatian, Czech, Danish, Dutch, English, Estonian, Faroese, Finnish, French, Galician, Georgian, German, Greek, Gujarati, Haitian Creole, Hausa, Hawaiian, Hebrew, Hindi, Hungarian, Icelandic, Indonesian, Irish, Italian, Japanese, Javanese, Kannada, Kazakh, Khmer, Korean, Lao, Latin, Latvian, Lingala, Lithuanian, Luxembourgish, Macedonian, Malagasy, Malay, Malayalam, Maltese, Māori, Marathi, Mongolian, Nepali, Norwegian, Nynorsk, Occitan, Pashto, Persian, Polish, Portuguese, Punjabi, Romanian, Russian, Sanskrit, Serbian, Shona, Sindhi, Sinhala, Slovak, Slovenian, Somali, Spanish, Sundanese, Swahili, Swedish, Tagalog, Tajik, Tamil, Tatar, Telugu, Thai, Tibetan, Turkish, Turkmen, Ukrainian, Urdu, Uzbek, Vietnamese, Welsh, Yiddish, Yoruba

Requirements

What you need.

LLT runs AI models on your Mac. This requires Apple Silicon and enough RAM for the models.

System

macOS 13 (Ventura) or newer. Apple Silicon (M1, M2, M3, M4 — any variant). 16 GB RAM recommended (Mistral 7B 4-bit needs ~4 GB, Whisper needs ~1 GB). Python 3.11+ for the backend.

Installation

DMG with the app + Install-Backend.command script. The script creates a Python venv, installs Whisper, MLX, Mistral, and all dependencies. Downloads models on first run (~4 GB total). One-time setup.

LLT DMG installer — drag to Applications

LLT Settings — Storage management for AI models

Simple DMG installation + model storage management

Pricing

One version. Everything included.

LLT — Local Live Translator for macOS

macOS · Apple Silicon

$29 one-time

🚀 Launch Price — $49 after launch

101 languages (Whisper + Mistral)
Auto language detection
Mic, app audio, or system audio
Floating translation overlay
Text-to-speech output
Transcript & SRT export
Auto-gain / manual gain
Menu bar app — always accessible
Multi-language interface (DE, EN, ES, FR, IT)
100% local — no cloud, no API keys
All future updates

📧 Get Notified at Launch

Requires Apple Silicon Mac (M1/M2/M3/M4) with 16 GB+ RAM. Backend installs automatically. Models download on first run (~4 GB).