Speak as anyone.
Live, in any app.
Change your voice in calls, streams and games with natural emotion. Type to speak in any voice, or brief an AI to talk for you in meetings. One API for developers.
Free plan · No card required · Works on Mac, Windows, Linux & Chromebook
Leo
ExpressiveHigh-energy, motivational · American
“Hey, it's me. This is what I sound like when I'm just talking normally on a call.”
Use it as your microphone in
Three ways to speak
Talk, type, or let it talk for you
Switch modes mid-call. Your voice, emotion and sounds carry across all of them.
Voice → Voice
Talk normally. Everyone else hears your chosen voice, word for word, a beat later.
- Starts speaking before you finish the sentence
- Keeps your words, swaps your voice
- Works as a microphone in any app
Type → Voice
Type a line and it plays instantly in your voice: for noisy rooms, accents or exact wording.
- Instant playback
- Quick-reply buttons
- Mix with live talking
Meeting agent
Brief it before a call. It listens, answers in your voice when addressed, and stays quiet otherwise.
- Answers only from your brief
- Stops when someone talks over it
- Approve replies before they play
Meeting agent
Brief it once. It handles the call.
Tell it who you are, what the meeting is about and what not to agree to. It listens, answers in your chosen voice when someone addresses you, and stays quiet when they don't.
- Answers only from your brief, and defers instead of inventing facts
- Whisper private instructions mid-call
- Approve mode: review every reply before it plays
Them
Ada, are we still on track for the November launch?
Agent · confident
Yes, we're on track for November 3rd. Design sign-off lands Friday.
Them
Could we squeeze in two more screens?
You whispered
hold the line on scope
Agent · calm
I hear you, but scope is locked for this release. Let's queue them for the next sprint.
Expression
Emotion on a button. Reactions on cue.
Set the mood of your voice live, or fire real recorded reactions: laughs, gasps, sighs and more, matched to your voice.
Emotions
Soundboard
Hotkeys work while you're in another app (desktop app). Upload your own sounds on paid plans.
Setup in minutes
How it works
Pick a voice
Choose from expressive presets, or clone your own with consent.
Select Vaux as your mic
One-time setup with a free virtual cable. Our guide covers every device.
Talk, type or brief
Switch modes, emotions and sounds live with one click or a hotkey.
For developers
The voice layer for character swaps
Already swapping faces in real time? Add the matching voice with a few lines. Short-lived browser tokens, server-enforced session limits, live controls and typed events.
- WebSocket streaming, ~0.5s to first audio
- Browser, Node & REST
- Per-minute metering & usage API
- Voices, clones, emotions, sounds
- Meeting agent endpoint
- Pairs with any video model
import { createVauxClient } from '@vaux/sdk';
const vaux = createVauxClient({
// Your backend mints a short-lived token; your API key never ships to the browser.
apiKeyProvider: () => fetch('/api/vaux-token', { method: 'POST' }).then((r) => r.json()),
});
const mic = await navigator.mediaDevices.getUserMedia({ audio: true });
const session = await vaux.realtime.connect(mic, {
voice: 'leo',
emotion: 'excited',
onAudioStream: (stream) => peer.addTrack(stream.getAudioTracks()[0], stream),
});
session.set({ emotion: 'calm' }); // change live
session.say('Typed lines play instantly.');
session.sound('laugh');
session.on('transcript', (t) => console.log(t.text));Who it's for
Built for real conversations
Streamers & creators
Characters, bits and personas on stream without a mod setup.
Privacy
Join calls and communities without exposing your real voice.
Accessibility
Type-to-speak in a natural voice that feels like yours.
Gaming & roleplay
Play the character, sound the part. Laugh, gasp and growl on cue.
Customer teams
A consistent brand voice across agents, shifts and accents.
Platforms
Pair with face/character-swap video for full identity swaps via API.
Pricing
Simple, usage-based plans
Credits cover live voice, type-to-speak and the agent. Top up anytime. Pay in dollars or naira.
Free
Try every voice
Free
100 credits / month
- ~8 minutes of live voice
- All preset voices
- Type-to-speak
- Soundboard
Creator
For streamers and callers
$9/month
1,200 credits / month
- ~100 minutes of live voice
- Ultra-low-latency voices
- 3 voice clones
- Meeting agent
- Conversation history
Pro
For power users and builders
$29/month
4,500 credits / month
- ~375 minutes of live voice
- 10 voice clones
- Meeting agent with saved briefs
- Priority support
Business
For teams and platforms
$99/month
18,000 credits / month
- ~1,500 minutes of live voice
- Premium expressive voices
- 50 voice clones
- Higher API limits
- Dedicated support
Consent-first cloning
Clones require proof the voice is yours or permitted. Clones stay private to your account.
Live audio not stored
Audio is processed in memory. Transcripts are kept only if you turn history on.
Zero tolerance for abuse
Impersonation for fraud, harassment or deception gets accounts closed and reported.
Read our Acceptable Use Policy and Voice Consent Policy.
FAQ
Questions, answered
Is it really real-time?+
Yes. It feels live: your new voice starts about half a second after you speak, and long sentences keep flowing while you talk. Your words stay exactly the same; only the voice changes.
Does it work in Zoom, Meet, Discord…?+
Yes. Vaux outputs to a virtual microphone (free to install), so any app that lets you choose a microphone can use it. Our tutorials walk through Mac, Windows, Linux and Chromebook.
Can I clone my own voice?+
Yes, on paid plans, and only with consent. You must own the voice or have written permission from the speaker. We verify consent and keep clones private to your account.
What about phones?+
Phones do not allow apps to replace the system microphone, so live use in phone apps is limited. You can use Type → Voice and the agent on mobile web, and join calls from a computer for full voice changing.
Is my audio stored?+
Live audio is processed in memory and not stored. Transcripts are saved only if you keep conversation history on, and you can delete them anytime. See our Privacy Policy.
Can developers use it?+
Yes. The same engine is available as an API and JavaScript SDK, built to pair with face and character-swap video models.