100% Offline · No Cloud · No Subscriptions
🛡️ Note: You may see a "Windows protected your PC" warning. Click "More info" → "Run anyway". We're working on getting a Microsoft licence.
🚀 BETA Working prototype — some glitches may occur. Report issues

Translate speech
in real time.

Voxxwire is a free, fully offline desktop translator that listens to your microphone and system audio, translates speech on the fly, and speaks the translation aloud — all without ever touching the internet.

Download for Windows Learn More
49+Languages
<1sLatency
0Data Sent
...Downloads
Voxxwire
YOU · ENG Hello, can everyone hear me?
JPN 皆さん、聞こえますか？
REMOTE · JPN はい、聞こえます。始めましょう。
ENG Yes, I can hear you. Let's begin.
Why Voxxwire?

Powerful Features,
Zero Compromises.

01

Fully Offline

All AI models run entirely on your machine. Your conversations never leave your computer — total privacy guaranteed.

02

Real-Time

Sub-second latency from speech detection to translated audio output, powered by Whisper ASR and Silero VAD.

03

Dual Channel

Translates your mic and system audio simultaneously — perfect for Zoom, Teams, and Google Meet calls.

04

49+ Languages

From Arabic to Vietnamese — translate between 49+ languages with auto-download of language packs on first use.

05

Live Subtitles

See real-time subtitles for both original and translated speech, with speaker labels and timestamps.

06

Free & Open

No subscriptions, no API keys, no hidden costs. Download once and use forever.

Live Preview

See It In Action

A real screenshot of the app — not a mockup. This is exactly what you'll see the moment you hit start.

Voxxwire app in use, showing a live translated conversation in the split You / Remote panels
01

Auto Device Detection

Picks your best mic and system audio output automatically — no manual setup.

02

Dual-Channel Split View

Your voice and the remote speaker's voice, translated side by side in real time.

03

Live Bilingual Subtitles

See the original and translated text together, with language tags per line.

04

One-Click Start

Pick your Whisper model size and hit start — you're translating in seconds.

Under the Hood

How It Works

01

Audio Capture

Captures your microphone and system audio streams in real time using low-latency audio drivers.

02

Voice Activity Detection

Silero VAD detects when someone is speaking and segments the audio into utterances, filtering silence.

03

Speech Recognition

OpenAI's Whisper model (via faster-whisper) transcribes the spoken audio into text with high accuracy.

04

Translation

Argos Translate converts the transcribed text from the source language to your target language, entirely offline.

05

Output

The translated text is displayed as live subtitles and spoken aloud through Piper TTS voices.

Supported Languages

Speak Your Language

🇬🇧

English

ASR & Translation

🇯🇵

Japanese

ASR & Translation

🇮🇳

Hindi

ASR & Translation

🇫🇷

French

ASR & Translation

🇩🇪

German

ASR & Translation

🇪🇸

Spanish

ASR & Translation

🇨🇳

Chinese

ASR & Translation

🇰🇷

Korean

ASR & Translation

🇸🇦

Arabic

ASR & Translation

...and 40+ more languages including Portuguese, Russian, Italian, Dutch, Turkish, Thai, Vietnamese, and more!

System Requirements

What You Need

✓ Operating System Windows 10 / 11 (64-bit)
✓ Memory 8 GB recommended (4 GB minimum)
~ Graphics Card NVIDIA recommended (CPU supported)
✓ Microphone Required for speech input
Voxxwire vs. Cloud Translators

Why Not Just Use The Cloud?

Most translation tools route your conversations through someone else's server. Voxxwire doesn't.

Voxxwire Cloud-Based Translators
Your audio ever leaves your PC Never Sent to a server
Works without internet Yes, fully offline Requires connection
Monthly cost $0, forever Often subscription-based
API keys / account required None Usually required
Latency <1s, local Network round-trip
Captures system audio (calls, meetings) Built in Rarely supported
Source code Open, on GitHub Closed
Advertisement
FAQ

Frequently Asked Questions

Is Voxxwire really free?

Yes — completely free and open source. No subscriptions, no API keys, no hidden costs. Download once and use it forever.

Does Voxxwire need an internet connection?

No. Speech recognition (Whisper), translation (Argos Translate), and text-to-speech (Piper) all run locally on your machine. The app only touches the network to download a language pack the first time you use a new language pair.

Is any of my audio or conversation data sent anywhere?

No. Everything is processed locally in memory. Nothing is uploaded, logged remotely, or shared with any third party.

Do I need an NVIDIA GPU?

No, a GPU is optional. Voxxwire runs on CPU too — a GPU just makes transcription and translation noticeably faster, especially with the larger Whisper models.

Why does Windows show a "protected your PC" warning?

Voxxwire doesn't yet have a paid code-signing certificate, so SmartScreen flags it by default. The app is fully open source — you're welcome to review the code on GitHub. Click "More info" → "Run anyway" to continue.

Which apps does Voxxwire work with?

Any app that plays audio through your system output — Zoom, Google Meet, Microsoft Teams, Discord, YouTube, and more — since Voxxwire captures system audio directly rather than integrating with a specific app.

Is Voxxwire open source?

Yes. The full source is on GitHub — issues, pull requests, and contributions are welcome.

Get Started

Download Voxxwire

Free. Offline. Forever yours.

Download for Windows (~210 MB)

Requires Windows 10/11 64-bit · ~850 MB after installation · No internet required after setup

or grab it straight from GitHub Releases

⚠️ SmartScreen Warning: When installing, Windows may show a "Windows protected your PC" message. Simply click "More info" → "Run anyway". This happens because we haven't purchased a code-signing certificate yet — the app is fully open-source and safe.