VoiceLayer FAQs

Question 1

How does VoiceLayer handle voice input (STT) and output (TTS)?

Accepted Answer

For voice output, it uses `edge-tts` for local Text-to-Speech, delivering spoken AI responses. For voice input, it utilizes `whisper.cpp` for local Speech-to-Text, transcribing your spoken commands quickly and privately.

Question 2

What are the different voice interaction modes available?

Accepted Answer

VoiceLayer offers five distinct interaction modes: `announce` for fire-and-forget TTS, `brief` for one-way explanations, `consult` to speak and hint at a response, `converse` for full Q&A with recording, and `think` for silent notes to a markdown log.

Question 3

Can I control when VoiceLayer stops recording my voice input?

Accepted Answer

Absolutely. You can manually stop recording by creating a specific touch file (`/tmp/voicelayer-stop`), or the system will automatically stop based on silence detection (e.g., 5 seconds of silence in converse mode) or a configurable timeout.

Question 4

Does VoiceLayer prevent microphone conflicts with other applications?

Accepted Answer

Yes, VoiceLayer features automatic session booking and lockfile management. This system ensures that only one instance uses the microphone at a time, preventing conflicts and ensuring a smooth voice interaction experience.

Question 5

What is VoiceLayer and how does it enhance AI coding assistants?

Accepted Answer

VoiceLayer integrates local voice input and output capabilities directly into AI coding assistant sessions (like Claude Code), enabling you to use spoken commands and receive AI responses audibly. It transforms your terminal into an interactive voice environment.

VoiceLayer

VoiceLayer

Key Features

Use Cases

Key Features

Use Cases