app-vox

app-vox / vox

Public

Your privacy-first voice-to-text tool. Local Whisper transcription with optional LLM enhancement so your audio never leaves your computer.

16
3
100% credibility
Found Feb 10, 2026 at 11 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

Vox is an open-source macOS menu bar app that locally transcribes speech to text using Whisper models, optionally enhances it with AI, and auto-pastes the result into the active application via customizable keyboard shortcuts.

How It Works

1
🔍 Discover Vox

You hear about Vox, a handy tool that turns your speech into text right on your Mac, and grab the latest version from the releases page.

2
📥 Install the app

Drag Vox to your Applications folder and launch it from the menu bar—it sits quietly ready to help.

3
⚙️ Download speech tool

In settings, pick and download a speech recognition pack—the small one is a great start for quick, accurate results.

4
🔓 Allow essentials

Grant microphone access to hear you and keyboard control to paste text automatically—Vox guides you smoothly.

5
Set your trigger
🤖
Add smart polish

Connect a helpful AI service to fix grammar and trim filler words for perfect text.

Skip to basics

Use raw speech-to-text right away without extras.

6
🗣️ Speak and paste

Hold your shortcut, talk naturally, watch the colorful indicator, and see clean text appear in any app like magic.

Effortless dictation

Now you dictate notes, emails, or code anywhere without typing—fast, private, and polished every time.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 11 to 16 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is vox?

Vox is a privacy-first voice-to-text app for macOS built in TypeScript with Electron and React. Hold a shortcut like Alt+Space, speak into your computer's audio input, and it transcribes locally using Whisper models while optionally polishing the text via an LLM like OpenAI or Bedrock—all without sending audio off-device. The result auto-pastes as Cmd+V into any app, solving clunky dictation tools that leak data to the cloud.

Why is it gaining traction?

Unlike cloud services, Vox keeps everything local with whisper.cpp for transcription and supports self-hosted LLMs, appealing to devs paranoid about voice vox github privacy. Features like hold/toggle modes, model size tradeoffs (tiny for speed, large for accuracy), and seamless tray integration make it a vox ai github workflow accelerator. Early buzz on github vox populi stems from its zero-config setup after downloading a ~460MB "small" model.

Who should use this?

Mac devs dictating code comments or notes during pair programming, writers brainstorming in Ulysses or Notes, or podcasters transcribing vox live stream intros locally. Ideal for remote workers avoiding vox tv programm cloud latency and Big Tech data grabs, especially with Bedrock or local LiteLLM proxies.

Verdict

Grab Vox if you're on Apple Silicon macOS and want offline voice-to-text; download the small model and test pasting transcripts today. At 10 stars and 1.0% credibility, it's raw but functional—docs are solid, tests exist, just needs Windows/Linux ports and more polish before production trust.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.