Voice Assistant Setup

Enable Voice Wake to turn OpenClaw into your hands-free AI assistant

Medium ⏱️ 10 min

What You’ll Get

Once configured, you can:

  • 🎙️ Wake with your voice — Say “Hey Claw” to activate
  • 🗣️ Natural conversation — Speak instead of typing
  • 🔊 Voice replies — AI responds with spoken audio
  • 🖼️ Live Canvas — Generate charts, code and more with voice commands

Prerequisites

  • OpenClaw installed on your Mac
  • macOS native app (voice features require the Mac app)
  • Microphone permissions granted

Step 1: Install the macOS App

If you haven’t installed the OpenClaw macOS app yet:

  1. Download the latest OpenClaw.app
  2. Drag OpenClaw.app to your Applications folder
  3. Right-click and select Open the first time to bypass Gatekeeper

Step 2: Grant Microphone Permission

The first time you use voice features, macOS will request microphone access:

  1. Open System Settings → Privacy & Security → Microphone
  2. Find OpenClaw and toggle it on
  3. Restart the OpenClaw app

Step 3: Configure Voice Wake

Enable voice activation in Terminal:

openclaw config set voicewake.enabled true
openclaw config set voicewake.keyword "Hey Claw"

You can customize the wake word:

# Use "Hey Assistant" as the wake word
openclaw config set voicewake.keyword "Hey Assistant"

Step 4: Enable Voice Replies

Turn on text-to-speech (TTS):

openclaw config set tts.enabled true
openclaw config set tts.voice "en-US-JennyNeural"

Available English voices:

Voice IDDescription
en-US-JennyNeuralFemale, friendly
en-US-GuyNeuralMale, professional
en-GB-SoniaNeuralBritish female
en-AU-NatashaNeuralAustralian female

Step 5: Start Using It

  1. Make sure the OpenClaw app is running in your menu bar
  2. Say the wake word: “Hey Claw”
  3. Wait for the activation chime, then speak your question
  4. The AI will respond with voice

Advanced Features

Continuous Conversation Mode

Enable continuous listening so you don’t need to say the wake word each time:

openclaw config set voicewake.continuous true

After the AI replies, it will automatically listen for your next question until you say “Done” or stay silent for 5 seconds.

Voice-Controlled Live Canvas

Use voice commands to generate visual content:

  • “Draw a bar chart showing this year’s sales data”
  • “Create a flowchart for the user registration process”
  • “Write Python code to implement quicksort”

The AI will render results in the Live Canvas window in real-time.

Voice Transcription

Save all voice conversations as text notes:

openclaw config set voicewake.transcribe true

All voice interactions will be automatically saved as text transcripts.

Troubleshooting

Wake Word Not Recognized

  • Try a different wake word with a more distinctive sound
  • Test in a quiet environment
  • Adjust microphone sensitivity settings

Voice Reply Delays

TTS generation requires an internet connection. If you’re experiencing latency:

  • Check your network connection
  • Try a different TTS provider
  • Consider local TTS (requires additional configuration)

System Says Microphone Access Denied

Make sure OpenClaw is authorized in System Settings → Privacy & Security → Microphone.

Next Steps