Hands-free AI voice assistant for drivers

Drive. Speak.
Everything onCue.

Your voice is the interface.

Get startedOpen source today  ·  Hosted product coming

How it works

Three steps.
That's the whole thing.

01

Dial with your voice

"Hey Siri, call OnCue." Any phone, any carrier, no internet — just a call. The voice agent you already have hands the rest to us.

02

Speak naturally

"Play something chill." "Skip this track." "Turn it up." OnCue understands you the way you actually talk.

03

The assistant does it

Actions happen instantly. You get a brief audio confirmation and you're back to the road.

What it controls today

Start with music.
The rest is coming.

Spotify is live in the open-source build. More integrations land as they're ready — each one has to work perfectly before it ships.

Live now

Spotify

“Play something to focus.”
“Skip.”
“Turn it down a bit.”
“What's this song?”

On the roadmap

  • Hosted app (no setup)Soon
  • Trip Planning2026
  • Food & Coffee Orders2026
  • Smart HomeLater

Why hands-free matters

Eyes on the road.
Always.

9×

more crashes when dialing manually vs. hands-free

5s

average eyes-off-road time to skip a track

88%

of drivers admit to phone use while driving

0

screen interactions required with OnCue

We're not here to lecture you about your phone. We've all glanced down at a red light. We've all half-reached for the aux cord. OnCue exists because the phone doesn't disappear — but the friction can.

Hands-free isn't a feature. It's a constraint that forces better design. Every interaction in OnCue was built to be done in a single spoken sentence, at 70mph, with one eye on a merge.

Your voice.
Not our data.

Trust is a precondition for hands-free. We built the privacy model first, then the product.

Calls are never recorded or used to train models. Audio is processed in real-time and discarded.

You control which services OnCue can access. Revoke any integration from your account page, instantly.

OnCue only acts on explicit commands. No passive listening, no background activity between calls.

FAQ

Questions.

OnCue is open source. Clone the repo, plug in your own keys for Twilio, Deepgram, ElevenLabs, Anthropic, and Spotify, and you have a working voice assistant in an afternoon. A hosted version with a full UI is in the works — until then, self-host is the way.
Self-hosting is free; you only pay the underlying providers for what you use. The hosted version will have a simple monthly plan when it launches.
Yes — any phone that can make a standard call. iPhone, Android, or even a basic handset. No app required. If you can dial a number, you can use OnCue.
The call ends gracefully. Any in-progress action is either completed or safely rolled back. When you reconnect, you just call again — no session to restore or resume.
Anytime. Hanging up is the equivalent of pressing stop. OnCue will finish any action it's already executing, then go idle.
Yes. Since OnCue is just a phone call, it plays through whatever audio output you already have connected — Bluetooth speakers, your car's system, a headset. It requires no special integrations with CarPlay or Android Auto.
English today. Spanish, French, and German are next on the roadmap.

Two ways in.
One available today.

OnCue is open source. Clone the repo, plug in your own keys, and you have a working voice assistant in an afternoon. A hosted version with a full UI is coming — sit tight or build now.

Read the setup guide

Self-host today  ·  Hosted product coming