Questions

Yes: your own Claude plan or Anthropic API key, billed by Anthropic. We never resell or include Claude access. Light use fits Claude Pro; a busy household may need more.

No. You sign in through Anthropic's own sign-in on the box. An API key stays on the box and is never shown or logged.

No. It's an independent product that runs Claude Code on your account.

Only for its name, on the device. Nothing is sent until you say it, reply right after it speaks, or talk in its first minute after start-up. Mute any time; it mutes itself overnight.

Your requests' words go to Claude on your account. Cloud speech, Google and remote access only if you turn them on. The full map.

Partly: face, board, reminders and speech work; thinking needs the internet.

A 64-bit Intel or AMD PC you can wipe (4 cores and 8 GB recommended), a screen and a USB speakerphone. Details.

Not yet.

Install is a USB stick and a wait; Claude sign-in takes minutes. Google and remote access take extra steps, and we help beta households.

It's a trade-off, so it gets its own. The AI can run commands there without asking today; a restricted permission mode is planned before public launch. Admin access starts off, and spending, messages, locks and alarms need your spoken yes.

Overnight, tested, with automatic rollback. Signed releases are being added in the beta; encrypted backups are planned.

Yes, but the owner must be an adult, and nobody is enrolled for face or voice recognition without consent.

English only, no smart-home control out of the box, no phone calls, no copy-protected music. It can be wrong; check what matters.

Free in the beta. Later the software stays free; optional Cloud Link is planned at $5/month (placeholder). Claude is separate.