EN ▾

Your Voice, Translated.
Anywhere. Private. Free.

The ultimate bidirectional Speech-to-Speech translator for VRChat, Discord, Zoom, and WhatsApp. Speak and hear back in real time. Runs 100% offline. Fits on a USB stick. No clouds, no subscriptions.

⬇ Download Portable v5.0 See All 28 Languages

*Free for personal use. Windows 10/11 + NVIDIA GPU required.

Why S2S Portable?

💾

USB Portable

No installation required. Unzip to a USB stick and run on any Windows PC. Your settings and models travel with you.

🔒

100% Privacy

What you say stays on your machine. All AI models run 100% offline on your hardware. No data is ever sent to the cloud.

🎎

Accurate Asian Languages

Specialized MeloTTS engine provides emotive, high-accuracy speech for Japanese, Korean, and Chinese.

🎙️

Voice Cloning

Don't sound like a robot. Record 20 seconds of your voice and the AI adapts the translation to match your unique timbre.

Impressive Local Speed

All AI runs locally on your NVIDIA GPU. Expect around 4-5 seconds on a good PC — and under 1 second on high-end RTX cards. No cloud needed!

🌍

28 Languages — Full GUI

The entire interface and all speech features work in 28 languages. The GUI adapts completely to your chosen language.

28 Languages, One App

Full speech-to-speech translation with a fully translated GUI. Every language works 100% offline.

Built for Everyone

S2S Portable bridges communication gaps. Whether you are mute, hard of hearing, or just prefer text — we have you covered.

⌨️ Text-to-Speech (Write to Speak)

Cannot speak? No problem. Type your message and the AI will translate and speak it out loud to your friends in VRChat or Discord.

📝 Speech-to-Text (Speak to Write)

Hard of hearing? The app provides real-time subtitles of what is being said and translated.

[SYSTEM] Initializing...
[USB]    Mode: Portable
[GPU]    NVIDIA RTX Detected
[AI]     Faster Whisper: Ready
[PRIVACY] Network: OFF
[STATUS] Ready to translate!
        

Under the Hood

Read the Manual