pricing · developer preview

Free while it's a preview.
Session time when it isn't.

Voqalize is early and we price it that way. Everything a developer needs to build a talking workflow and put it in front of real users costs nothing today. When production pricing arrives it will meter one thing — the time a session is live on our voice plane — and we'll publish the rates before a single session is billed.

statement · developer preview
voice planeWebRTC, turn detection, interruption, recordingincluded
speechSTT + TTS, models we run ourselvesincluded
sdksPython, React — source on GitHubincluded
dev environmenttenant, agents, publishable keysincluded
demoseleven of them, running, source and allincluded
supportyou talk to the people who wrote itincluded
due today$0

Not a trial with a clock on it. The preview is how the product is sold right now, and this is the whole statement.

when it's priced

One meter. And a bill you can predict.

We'd rather tell you the shape now than surprise you with it later. Four things are already decided, because they fall out of how Voqalize is built.

The unit is session time

You pay for the time a session is live on our voice plane. Not seats, not agents, not requests.

Your model bill stays yours

We never resell tokens, because we never buy them. Your LLM runs in your backend and bills you directly — that's the whole architecture, and it's why this line can't appear on our invoice.

Speech is not a passthrough

STT and TTS are our own models on our own hardware, so speech is part of the session price rather than a third-party voice vendor marked up.

Rates land before charges do

We'll publish the numbers and tell you before anything meters. No preview session turns into an invoice after the fact.

the short answers

How do I get in?
Create an account, take a publishable key, and go. Nothing on the developer path is gated behind a call.
What are the preview limits?
Enough concurrent sessions to build and demo. If you need more for a pilot, ask — we raise it by hand today.
Can I run this in my own VPC?
The speech models are built to deploy inside your network. Talk to us about what that looks like for your setup.
What happens to my agents when pricing starts?
They keep running. You get the rates in writing first, and a window to decide.

Start without asking anyone: the brain guide gets an agent talking, and the demo gallery shows what one looks like finished.

Tell us what you want to make talk.

Bring a workflow and a rough sense of volume. We'll tell you straight whether this is ready for it.

hello@voqalize.com

Try the order workflowlive app · real session