Analysis updated 2026-05-18
Follow a foreign-language video call or lecture with live translated captions
Handle two-way bilingual conversations with automatic language detection
Have translated text read aloud using text-to-speech
Run fully offline on Apple Silicon using local models
| phuc-nt/my-translator | getify/monio | rebel0789/codexpro | |
|---|---|---|---|
| Stars | 1,106 | 1,126 | 1,134 |
| Language | JavaScript | JavaScript | JavaScript |
| Last pushed | — | — | 2026-06-30 |
| Maintenance | — | — | Active |
| Setup difficulty | moderate | moderate | moderate |
| Complexity | 3/5 | 4/5 | 3/5 |
| Audience | general | developer | developer |
Figures from each repo's GitHub metadata at analysis time.
Requires your own API keys for Soniox and optionally Google Cloud or ElevenLabs.
My Translator is a desktop app that listens to audio, from your microphone, your system's speakers, or both, and translates speech in near real time, displaying the translated text in a small floating overlay on your screen. It targets people who need to follow conversations, video calls, lectures, or media in a language they don't fully understand. The app captures audio at the system level, sends it to a third-party speech-to-text and translation API called Soniox, and gets back a translated transcript with roughly two to three seconds of delay. It supports translation between over 70 source languages. Optionally, it can also read the translated text aloud using one of three text-to-speech providers: a free built-in option, Google Cloud, or ElevenLabs. Two-way translation mode lets you handle bilingual conversations, the app detects which language is being spoken and translates to the other language automatically. Privacy is a key design principle: there is no relay server. The app connects directly to whatever API services you configure using your own API keys, stores those keys locally on your machine, and collects no usage data or analytics. An experimental offline mode is available for Apple Silicon devices only, using local models to run entirely on-device. The app is built with Tauri, a framework that uses a Rust backend combined with a web-based frontend, and runs on macOS and Windows. The project is open source under the MIT license.
A desktop app that listens to audio and shows near real-time translated speech in a floating overlay on your screen.
Mainly JavaScript. The stack also includes Tauri, Rust, JavaScript.
Use freely for any purpose, including commercial use, as long as you keep the copyright notice, under the MIT license.
Setup difficulty is rated moderate, with roughly 30min to a first successful run.
Mainly general.
This repo across BitVibe Labs
Verify against the repo before relying on details.