01
Full voice conversation.
Talk to it the way you'd talk to anyone. It listens, understands, and talks back through a real speech pipeline.
voice
a desk creature · by presence labs
Meet Sudo.
A small, open-source creature for your desk. It talks back, it has opinions, and it runs on whatever AI you point it at. Yours to shape. No subscription.
Not a gadget. A creature you can talk to, shape, and call your own.
talks backremembers youyours to shapeno subscription
02 · meet sudo
The fastest way to explain it is to watch it move.
Face, voice, moods, memory, and the part where it quietly judges your setup. A video does that better than another wall of text.
launch film
Sudo in motion.
Face, voice, moods, memory — the whole thing in one pass. Hit play.
01face and moods
02runs dry
03notices you're there
04remembers locally
05talks back
03 · capabilities
Talk to it. Shape it. Build on it.
Three things make Sudo more than a gadget: a real voice pipeline, a personality you can open and tune, and a mods ecosystem that turns it into whatever you want next.
02
An open personality to tweak.
The character is yours to shape. Adjust how it reacts, what it notices, and how it talks to you.
personality
03
A mods ecosystem.
Control your Spotify, set focus sessions, wire it to your tools, and build whatever you imagine on top.
mods
04 · run it, build on it
We built the hard part. You build everything on top.
Sudo isn't a black box with a face on it. The AI runs where you put it, the SDK is real, and the personality engine is the part we actually nailed.
01
Run it your way.
Self-host the AI on your own machine, point it at a free API tier, or use our cloud. No subscription unless you want one. Your data stays where you put it.
our cloud
free for early adopters, low cost later. zero setup.
any free API
point Sudo at a free tier. pay nothing.
self-host
run a model on your own machine. pay nothing, permanently.
Same device on all three paths. Nothing behind a paywall.
02
Build on it.
A real SDK with docs and examples. Write behaviors, wire it to your tools, and ship whole new characters on top of the runtime.
behaviorsreact to events, drive moods and gestures
integrationswire Sudo to your commits, your music, your terminal
charactersauthor and ship a whole new personality on the runtime
03
The engine is ours.
The face runs on a circumplex emotion model. Every expression is a point in valence and arousal space, and the creature blends toward a live target frame by frame, with a liveness layer adding breathing, sway, and gaze drift on top. The SDK exposes that target plus the inputs and outputs. The engine stays ours. Everything you'd build on it is yours.
// sudo.behavior.ts
// nudge Sudo toward "curious" when a build starts
sudo.on("build:start", () => {
sudo.emotion.target({ valence: 0.3, arousal: 0.7 });
sudo.gesture("curious_tilt");
});Nothing you build depends on us shipping it for you. That's the point of an SDK.
06 · founder's edition
Be in the first batch.
The first run is twenty numbered units. The earliest way to get a Sudo, by application.
by application20 units
01/20
- twenty units, individually numbered.
- lifetime cloud AI, free for the life of the unit.
- a direct line to the founder, not a support queue.
- by application. tell us why you want in, we read every one.
Retail pricing will be lower than Founder's Edition. We'd rather say that now than burn the people who backed it first.
07 · faq
Questions that should have plain answers.
No over-claiming. If something's ours, we say so. If something's open, we say exactly what that means.
+Do I need to be a developer to use it?
No. Out of the box, Sudo works immediately. It talks, learns your routine, reacts, and keeps you company. The SDK is there for people who want to build on it. The default experience needs zero technical skill.
+Is Sudo open source?
The SDK is open, with docs and examples. The firmware and the personality engine are ours. What's open is what you build on; what's yours is your data and your AI endpoint. We're specific about this on purpose.
+How is this different from EMO?
Three ways. Price: we're meaningfully cheaper. AI: you can run Sudo on your own endpoint or a free API, EMO can't. Building: we ship a real SDK. You get control over the experience, even though the engine itself is ours.
+How is this different from Alexa or Google Home?
Those are utilities that live in someone else's cloud. Sudo is a creature that lives on your desk, with a personality that develops from how you treat it, and an SDK you can build on. It's not a voice assistant. It's a creature you can extend.
+What's the real monthly cost?
Zero, if you choose. Use a free API tier or self-host and you pay nothing. Use our cloud and it's free for early adopters, low cost later. We don't gate features behind a subscription.
+What about my data?
Memory is stored locally, and cloud calls send only the immediate context, not your history. If you self-host the AI, your interactions never leave your machine. You choose where the intelligence runs.