Voice Assistant Overview

Voice Assistant Overview

The on-device voice assistant runs supported wake-up and command handling on the wearable. Applications can read the device capability, inspect its current mode, and request a compatible mode through OnDeviceVoiceAssistantAPI.

Available Features

Browse the generated Wiki navigation for the pages in this section.

Capability and Mode

OnDeviceVoiceAssistantCapability describes what the device supports:

Capability	Supported behavior
.none	No on-device voice-command capability.
.keywordWakeup	Keyword wake-up only, such as “Hi Buds.”
.common	Common commands such as playback, track, and volume control.

OnDeviceVoiceAssistantMode describes the active behavior:

Mode	Behavior
.unknown	Current mode is unavailable. Do not set this value.
.disabled	On-device voice assistant is disabled.
.basic	Wake-up and basic voice commands.
.advanced	Basic behavior plus call-state voice control.

Capability determines which modes should be offered. The current Demo exposes disabled and basic for .keywordWakeup, and disabled, basic, and advanced for .common.

Integration Flow

Voice assistant integration path

Prepare the SDK integration, confirm authorization and device capability, then expose only compatible assistant modes.

1. Initialize SDK — Install and initialize the base AIBuds SDK for the host application.
2. Install Integration — Add AIBudsVoiceAssistant when service authentication is required.
3. Complete Authorization — Finish the required voice-assistant service authorization flow.
4. Connect Device — Verify OnDeviceVoiceAssistantAPI conformance on the active device.
5. Read Capability & Mode — Inspect supported behavior and the currently active assistant mode.
6. Apply Compatible Mode — Offer only valid modes, apply the selection, and update UI from completion or change callbacks.

Prerequisites

• The device is connected and ready.
• Required service authentication is complete.
• The device exposes OnDeviceVoiceAssistantAPI and a capability other than .none.