OPEN SOURCE ALL-WAY VOICE LAYER

Break out of the text box.

VoiceFi goes beyond one-way dictation to keep you in effortless focus, giving your agents a voice when they finish and letting you steer, verify, and delegate across Claude Code, Google Antigravity, Gemini, Cursor, Codex, and more without typing a key or switching windows. And with Remote Companion, you can do it all from your phone.

$ brew install atxatlarge-code/tap/vifi
Get VoiceFi for Mac → View Docs & CLI →
✓ 0 Screen Permissions (Audio-only IPC)
✓ 100% Local Apple Neural Engine (0 Cloud)
✓ Zero Sudo / Root Required (Userland install)
✓ Free & Open Source (MIT on GitHub)

Choose Your Character

Viv

VoiceFi

Christopher

Emily

Aria

Sonia

Code Reviewer

Stefan

ProActive Feedback Loop

Use your voice to guide your agents

VoiceFi Mobile Headset Companion
9:30
5G
VoiceFi

VoiceFi Mobile

STANDBY
Ready to Listen
TRANSCRIPT STREAM 00:00
“Agent finished unit tests. Say 'Deploy to staging' to dispatch...”
HUD macOS
VoiceFi Ready (⇧⌘N)
Standing by • Dictate (⌃T) or speak to agent (⌃R)
Terminal

VoiceFi Core v0.8.4 (darwin-arm64 • Apple Metal)

Local bridge listening on /tmp/voicefi.sock (127.0.0.1:5141)

Ready for agent events. Press 'Start Live Loop' to test...

Local is Faster

Time from spoken thought to IDE code execution. Metal on-device compute vs. network transit.

Local Network Cloud Friction
VoiceFi Local Pipeline
120ms ⚡ Instant
Superwhisper / Local
800ms 6.7× Slower
Cloud Speech Pipelines
1,450ms 12× Slower
Manual Alt-Tab / Dictation
2,800ms+ 23× Slower
Agentic Differentiators

One-Way Dictation
vs.
The VoiceFi (vifi) Feedback Lp

Traditional dictation tools only replace typing. VoiceFi connects the entire circular feedback loop with your agents, tools (via MCP) and anything that has terminal access.

← Swipe to compare all platforms →
Capability VoiceFi (vifi) Wispr Flow Superwhisper / MacWhisper ElevenLabs Cloud
Two-Way Feedback Loop Agent speaks turn soundbite → Auto-listens → Developer replies ✓ Continuous 2-Way Loop ✗ 1-Way Dictation only ✗ 1-Way Dictation only ✗ Disconnected from IDE
AI Agent Hook Integration Native stop & notification hooks for Antigravity, Claude Code, Cursor, Aider ✓ Native 1-Line Setup ✗ No Agent Hooks ✗ No Agent Hooks ✗ No IDE Agent Hooks
Zero-Focus Agent Dispatch & IPC Dispatches tasks & prompts across any terminal or agent via native IPC sockets — 0 focus loss, 0 clipboard clobbering, 0 screen flicker ✓ Native Background IPC (vifi send) ✗ Foreground paste only (steals active window focus) ✗ Simulated keystrokes (steals focus & overwrites clipboard) ✗ Disconnected from local terminals & IDE sockets
Universal Voice MCP Server Model Context Protocol support — allows any AI agent or tool to talk, listen & synthesize ✓ Native Voice MCP (Any tool can talk) ⚠️ Read-Only Notetaker MCP (No speech tools) ✗ No MCP Support ✗ Proprietary API Only (No MCP)
"Second Voice" & Obsidian Vault Plugin Dual-voice layer for your Second Brain — ties multi-agent vaults to Obsidian via native plugin & MCP ✓ Multi-Voice & Native Obsidian Plugin ⚠️ Meeting transcripts only (No voice plugin/TTS) ✗ No Obsidian Plugin or Multi-Voice ✗ Disconnected from Vaults
Hands-Free Auto-Listening (VAD) Mic arms automatically when agent finishes a turn (zero clicking) ✓ Automatic Turn-Based VAD ✗ Push-to-Talk only ✗ Push-to-Talk only ⚠️ Continuous Cloud Stream
Smart Turn Summarization Cleanses raw terminal logs into 1-sentence spoken briefings ✓ Intelligent TTS Briefing ✗ No TTS / Readout ✗ No TTS / Readout ✗ Generic Voice Output
Audio Privacy & Egress Local Metal/CoreML Whisper in RAM vs cloud audio streaming 🔒 100% On-Device (0 Egress) ⚠️ Cloud Audio Upload 🔒 0 Bytes Egress (Offline) ⚠️ Cloud Audio Streaming
Cost & Licensing Open source CLI freedom vs monthly recurring software subscriptions 100% Free MIT CLI $12 – $20 / month $20 – $30 one-time / sub $0.05 – $0.30 / min API
Looking for deeper technical & latency specs? Explore full stage-by-stage latency breakdowns, Unix socket IPC benchmarks, and runtime matrices.
Full Architecture & Benchmark Matrix →
Universal Compatibility & Protocols

Agent & IDE Ecosystem

Zero-latency voice and acoustic layer connecting autonomous coding agents, second-brain vaults, and system terminals.

DRAG OR TAP CONNECTOR LINES TO PLUCK ACOUSTIC HARP
VoiceFi Core VoiceFi PORT 5141 CORE SERVER
1. Autonomous Agents

Antigravity

Subagent Hooks

AUTO

Claude

Terminal REPL

STDIO

Gemini

Live & Flash SDK

VOICE

Cursor

Voice Dispatch

HOTKEY
2. Knowledge Vaults & Editors

Obsidian

Second Brain

PLUGIN

VS Code

Extension

IDE
3. Protocols & Shell

MCP

Model Context Protocol

STDIO

WebMCP

navigator.modelContext

W3C

Ollama

Local LLMs

OFFLINE

Terminal

Ghostty • Warp • iTerm • tmux

PIPES
Google Antigravity AUTO-HOOKED

Transport: Unix domain socket (/tmp/voicefi.sock) • Hook: agent.turn_completed

vifi setup --dev