Your typing becomes a real mic in Discord and any game.
Type a phrase, hear a natural voice, and it comes out of your virtual microphone — so everyone in your call or game voice chat hears it. Hundreds of voices work instantly, and a local voice takes over automatically when you go offline.
A voice that shows up where you actually talk.
Built for people who can't or prefer not to speak, for gamers who want a voice in chat, and for anyone who needs fast, private, assistive communication.
Virtual mic for Discord & games
The generated voice routes into a virtual microphone (VB-Cable, bundled in the installer). Any app that reads a mic hears it — Discord, Zoom, in-game voice chat.
Smart Auto routing
Online Edge TTS when you have internet, instant fallback to a local voice (Kokoro/Piper) when you're offline or the connection drops. No setting to flip.
Instant, no account
Hundreds of natural Microsoft Edge voices work the moment you open the app. Nothing to download, no sign-up, no telemetry.
Global voice shortcuts
Bind a phrase to a hotkey and fire it from inside a game without alt-tabbing. "GG", "cover me", or your own lines — one keypress.
Built for accessibility
Persistent history and draft, quick phrases, an on-screen keyboard, high-contrast and large-font modes, full keyboard navigation, visible focus.
Free & open source
MIT licensed, no paywall, no data collection. Local synthesis never leaves your machine. Contributions welcome.
Type → route → speak into any voice chat.
Auto mode picks the best engine per phrase and fans the audio to your virtual mic so Discord and games hear it — while you still hear yourself.
Cloud quality by default, local when you need it.
Auto mode selects the local engine by your hardware. The recommended local path is Piper first, Kokoro for higher quality.
| Engine | Status | Notes |
|---|---|---|
| Edge TTS (cloud) | Default | Hundreds of voices, no install, needs internet |
| Piper | Stable · local | Lightweight, CPU-only, works offline |
| Kokoro | Stable · local | Higher quality (MOS 4.2), still CPU-friendly |
| XTTS v2 | Advanced | Voice cloning; NVIDIA + CUDA only |
| MeloTTS · Fish Speech · Bark | Experimental | Outside the main flow |
Running in three steps.
Windows 10/11 x64. The virtual microphone is bundled and sets itself up.
Download
Grab the latest VoiceLaunch-TTS-Setup.exe from GitHub Releases and run it. VB-Cable installs automatically.
Enable the mic
Open the app, turn on the virtual mic (it auto-selects CABLE Input), type a phrase, and hit Speak.
Point Discord at it
In Discord (or any game) set the input device to CABLE Output. Everyone in the channel now hears your voice.
Get-FileHash -Algorithm SHA256 .\VoiceLaunch-TTS-Setup.exe
A voice station, not just a toy.
Designed with real assistive use in mind — every state combines color, icon, and label, and nothing depends on color alone.
Give yourself a voice in every call.
Free, open source, no account. Download it, enable the mic, and you're talking in Discord in under a minute.