Real-time RVC engine
Sub-second mouth-to-ear voice transformation, locally on your GPU. Auto-adjusts under load.
download
Get PhazeVox.
Free, local, and yours.
Free to download. Free starter voices included. The full premium catalog is one click away inside the app once you have a Studio subscription.
Local-only audio · no telemetry · cancel Studio anytime
Heavier than a chat app, lighter than a game. The installer ships the full on-device AI engine — your microphone never leaves your machine, so all the model weights, runtime, and DirectML GPU compute live in the install instead of on our servers.
premium roster
Plus a cast you'll actually want to use.
Free starter voices ship with the install. The character roster below downloads on-demand inside the app once you've got Studio.
in the box
Everything you need to start.
Starter voice catalog
Six free starter characters + DSP effects (Telephone, Robot). No subscription required.
Studio-ready
Once you have a Studio subscription, the full premium roster downloads on demand from inside the app.
what it looks like
A real-time console, not a setup wizard.
Pick a voice, hit Start, talk into your mic. The Console handles the rest — latency monitoring, audio routing, push-to-talk. No tweaking required to be on air.
install & setup
Three steps. No tweaking required.
Run the installer
Double-click the .exe. Windows SmartScreen may prompt on early builds — click "More info" → "Run anyway."
Install the virtual audio device
On first launch PhazeVox prompts to install VB-Cable — a free virtual audio device by VB-Audio Software that lets Discord, Steam, and OBS see PhazeVox's output as a microphone. One click; we handle the rest.
Choose a voice and hit start
Open the Library, click any starter voice, then press Start in the Console. Your friends will hear the transformed audio.
system requirements
What you'll need.
Minimum
- Operating system
- Windows 10 / 11 (64-bit)
- GPU
- DirectML-capable (Intel / AMD / NVIDIA)
- CPU
- Modern 4-core (last ~5 years)
- RAM
- 8 GB
- Disk space
- ~2 GB
- Audio
- Any microphone Windows recognises
Recommended
- Operating system
- Windows 11 (64-bit)
- GPU
- NVIDIA RTX 3060 or better
- CPU
- 6-core+ from the last 3 years
- RAM
- 16 GB
- Disk space
- ~4 GB (full premium catalog)
- Audio
- USB or XLR mic with a clean signal
No dedicated GPU? PhazeVox still installs and runs, but expect noticeable latency (~150-250 ms) and occasional stutters on AI character voices. The DSP effects + new broadcast mic presets (Clean Studio Mic, Podcast Voice) run smoothly on any modern machine.
Mac and Linux aren't supported at launch. We're focused on shipping the Windows build first; if there's enough interest after launch we'll evaluate other platforms.