> Voice Mode
Set up push-to-talk voice commands with on-device STT and ElevenLabs TTS.
Enabling Voice Mode
PRO Voice mode uses on-device speech recognition (expo-speech-recognition) for STT and ElevenLabs for TTS, proxied through Supabase Edge Functions.
- On the mobile app, go to Settings → Voice Mode → Enable.
- Grant microphone permissions when prompted on mobile.
Push-to-Talk Usage
Hold the microphone button on the mobile app to speak your command. Release to send the transcribed text to Claude.
Speech is transcribed on-device using the platform's native speech recognition engine. The resulting text command is then processed by Claude on the Morpheus Agent exactly as if you had typed it. Responses are spoken aloud via ElevenLabs TTS.
Strict Voice Mode
When strict voice mode is enabled, the text input field is disabled. Only voice commands are accepted. This is useful for hands-free scenarios where you want to avoid accidental typed input.
Enable it in Settings → Voice Mode → Strict Mode. Toggle it off at any time to restore the text input.
Audio Troubleshooting
Microphone not working
Check that the mobile app has microphone permissions. On iOS, go to Settings → Morpheus → Microphone. On Android, go to Settings → Apps → Morpheus → Permissions.
Transcription not responding
Ensure the mobile device has an active internet connection for ElevenLabs TTS. Check that speech recognition permissions are granted in your device settings.
High latency or dropped audio
Voice commands require a stable network connection. If using a Cloudflare Tunnel, latency may be higher. For best results, use LAN mode.