AuraLang

Why I built this

Millions of people consume content in languages they don't master and lose context every time they pause to translate. I wanted to solve that: real-time audio translation, with no API keys or backend of my own. Transcription happens on-device with Whisper — only the transcribed text goes over the network to be translated.

What it does

• Any tab with audio — videos, calls, podcasts, live streams.
• On-device transcription — Whisper runs locally; the audio never leaves your machine.
• Spoken translation — the original audio is muted; you only hear the translation.
• Pick source and target language, and switch pairs whenever you want.

How it works

1. 01

   Select the audio from any tab and choose the source and target language.

2. 02

   The first time, the Whisper model is downloaded to the device.

3. 03

   Once downloaded, the audio is transcribed in real time and translated on-device.

4. 04

   The translation plays on the device and the original audio is muted.

Usage modes

Privacy

No backend, no telemetry, no account. Transcription happens on your device; only the transcribed text is sent to Google Translate to get the translation. Settings (languages, theme) are stored in chrome.storage.local on your machine.

Build & Tooling

Links