OpenClawSkills
GitHub
Feature • TutorialHeader.readTime

Voice Features

Speak naturally to your agent. Always-on wake word detection and human-like speech synthesis.

OpenClaw supports voice interaction through voice wake and talk mode, allowing you to interact naturally using speech. Voice features are available on macOS, iOS, and Android platforms.

Tutorial.step

Voice Wake

<title> skill enables always-on voice recognition:

  • Always Listening - Continuously monitors for wake words
  • Hands-Free - Activate OpenClaw without typing
  • macOS/iOS/Android - Available on supported platforms
  • Privacy-Focused - Processing happens locally when possible

macOS Voice Wake

On macOS, voice wake integrates with the system:

  • Uses macOS speech recognition
  • Accessible via menu bar app
  • Configurable wake words
  • Background processing

Voice Wake Configuration

VoicePage.wake.configDesc

voicewake.json
{
    "triggers": ["hey openclaw", "computer"],
    "updatedAtMs": 1730000000000
  }
Tutorial.step

Talk Mode

Talk mode provides two-way voice conversation:

Speech Input

Speak your requests naturally.

Audio Output

Hear responses as speech.

ElevenLabs Integration

High-quality text-to-speech.

Natural Conversation

Feels like talking to a person.

Talk Mode (TTS) Configuration

VoicePage.talk.ttsDesc

openclaw.json
{
  "messages": {
    "tts": {
      "auto": "always",
      "provider": "elevenlabs",
      "elevenlabs": {
        "voiceId": "pMsXgVXv3BLzUgSXRplE"
      }
    }
  }
}
Tutorial.step

Platform Support

  • macOSFull voice wake and talk mode support.
  • iOSVoice features via iOS node.
  • AndroidVoice features via Android node.
Tutorial.step

Audio Handling

OpenClaw handles audio in multiple ways:

  • 🎙️
    Voice NotesSend and receive voice messages.
  • 📝
    TranscriptionAutomatic transcription of audio.
  • ⚙️
    Audio ProcessingExtract information from audio.
  • 📻
    Media PipelineHandles audio files efficiently.

Transcription Hooks

Configure transcription hooks to process voice notes:

  • Automatic transcription of incoming audio
  • Process transcribed text as regular messages
  • Support for multiple transcription services
  • Configurable per channel
Tutorial.step

Use Cases

  • 1
    Hands-Free OperationControl OpenClaw while driving or cooking.
  • 2
    Quick InteractionsAsk questions without typing.
  • 3
    AccessibilityVoice interaction for users who prefer speech.
  • 4
    MultitaskingInteract while doing other tasks.
Tutorial.step

Platform-Specific Features

macOS
  • Menu bar integration
  • System-wide voice wake
  • Native speech recognition
  • Background processing
iOS
  • Voice features via iOS node
  • iPhone/iPad support
  • Native iOS speech APIs
Android
  • Voice features via Android node
  • Android speech recognition
  • Background voice processing