Cloud AI, local AI, and what happens offline
A plain-language guide to latency, privacy, subscriptions, updates, and the support horizon behind a robot’s personality.
Products, prices, policies, and evidence change. Verify the current primary sources for the exact model and region before acting.
Most robots are hybrid systems
Local processors usually handle balance, obstacle avoidance, wake words, touch responses, and basic animation. Remote services may handle speech recognition, large language models, long-term memories, content updates, and account features.
That split matters more than a simple online or offline label. Look for a feature-by-feature description of what survives when Wi-Fi is unavailable.
Cloud dependence changes ownership
Cloud features can improve conversation and evolve quickly, but they also create recurring cost, service continuity, and data-governance questions. Local features tend to be more durable and private, but may be narrower or update less often.
Run a feature-by-feature offline test
Disconnect Wi-Fi during a return window and test waking, movement, touch responses, charging, reminders, saved routines, speech, and the companion app separately. Record degraded behavior rather than reducing the result to “works” or “does not work.”
Ask whether local functions require a recent cloud token or account check. A robot can appear offline-capable during a short outage yet still stop after credentials expire or a service is retired.
What works immediately without Wi-Fi?
What still works after a day?
Can local data be backed up without the vendor cloud?
Does the product disclose a service end-of-life plan?
Framework sources
These sources shape the questions in this guide. They do not certify any listed robot.