Talk to your
agents.

You type 40 words per minute. You speak 150. VoiceLayer adds voice to Claude Code, Cursor, and every MCP client. Press F6. Speak. Ship.

free · open source · local-first · no cloud

Get started View source
Recording
Transcribed · whisper.cpp · 1.2s

Split the JWT validation into its own middleware function. Add refresh token rotation and keep the cookie fallback for the admin dashboard.

<1.5stranscription
150 WPMspeech vs 40 typing
HE + ENbilingual
On-devicewhisper.cpp
The bottleneck

Your fingers are the slowest part

You think at the speed of speech. Your AI should listen.

Without VoiceLayer
  • ×Type every instruction at 40 WPM
  • ×Hands glued to keyboard during QA
  • ×Context-switch to type quick notes
  • ×150 WPM brain, 40 WPM fingers
  • ×No way to do voice-first workflows
With VoiceLayer
  • &check;Speak naturally to your AI at 150 WPM
  • &check;Hands-free code review and QA testing
  • &check;Voice notes while browsing the app
  • &check;Local transcription in under 1.5 seconds
  • &check;5 voice modes: announce, brief, consult, converse, think

Prefer typing? VoiceLayer is optional per-session. Enable it when you want speed. Disable it in shared offices. Your workflow, your choice.

Capabilities

Built for devs who think faster than they type

On-device STT

whisper.cpp large-v3-turbo runs entirely on your Mac. Sub-1.5s transcription, Hebrew + English in the same sentence. No cloud, no API key, no latency.

VoiceBar

Native SwiftUI menu bar app. F6 push-to-talk via Karabiner, live waveform pill, real-time teleprompter. Collapses to a dot when idle. Draggable anywhere.

MCP tools

voice_speak for non-blocking TTS with auto mode detection (announce, brief, consult, think). voice_ask speaks a question aloud, records your voice response via Silero VAD, and returns the transcription.

How it works

F6 → text at your cursor in <2 seconds

F6
Hold F6
0ms
Karabiner hotkey triggers VoiceBar recording
Speak
~3s
Silero VAD detects speech & silence
Transcribe
~1.1s
whisper.cpp large-v3-turbo, on-device
Paste
instant
Auto-pastes into your active field
Works with

Any MCP client

Claude Code
Claude Code
Cursor
Cursor
Zed
Zed
VS Code
VS Code
Codex
Codex
Kiro
Kiro
Gemini CLI
Gemini CLI
Getting started

Three steps

01

Install from npm

02

Launch VoiceBar

03

Press F6 and start talking

Stop typing. Start talking.

One install. On-device STT. Works with any MCP client.

Companion App

Meet VoiceBar

A native SwiftUI macOS menu bar app that gives VoiceLayer a persistent visual presence. Always on, always ready.

F6 push-to-talk

Hold F6 to record, release to transcribe. Global hotkey via Karabiner.

Live teleprompter

Words appear as TTS speaks. Auto-scroll, word-level highlighting.

Floating pill UI

Waveform visualization during recording. Collapses to a dot when idle. Draggable anywhere.

Requires VoiceLayer MCP server. Install with voicelayer bar