ZhangHanDong

ZhangHanDong / vox

Public

Voice input reimagined — speak in any language, type in any language.

93
11
100% credibility
Found Apr 03, 2026 at 85 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Rust
AI Summary

Vox is a macOS menu bar app for press-to-talk voice dictation that transcribes speech locally and optionally refines or translates text before inserting it at the cursor.

How It Works

1
🔍 Discover Vox

You find Vox, a handy Mac app that lets you speak to type text anywhere on your computer.

2
⚙️ Prepare Voice Listener

You start a simple voice service on your Mac so it can understand your speech locally.

3
🚀 Launch the App

Click to run Vox, and it appears quietly in your menu bar as a little mic icon.

4
✅ Grant Permissions

Your Mac asks for microphone and keyboard access, so you say yes to let it work smoothly.

5
🎙️ Speak and Type

Hold the Option key, talk naturally in your language, release – and your words appear right where your cursor blinks!

6
🌐 Pick Language or Polish

From the menu, choose your language or turn on smart fixes to translate or correct as needed.

✨ Effortless Typing

Now you dictate emails, notes, or code in Chinese, English, or more, feeling like magic every time.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 85 to 93 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 Rust-powered macOS menu bar app for seamless voice input—hold Option, speak in Chinese, English, Japanese, Korean, or more, release to transcribe and paste text anywhere your cursor sits. Built with Makepad UI and local Qwen3-ASR via OminiX-API, it handles 30+ languages with optional LLM refinement for error fixes, real-time translation (e.g., speak Chinese, type English), or classical Chinese conversion. Privacy-focused: all audio stays local on Apple Silicon, no cloud leaks.

Why is it gaining traction?

It outshines stock voice input mac or voice input keyboard tools with accurate local ASR (CER 5.88 on Chinese) and smart post-processing, skipping github voice change gimmicks or github voice cloning ai hype for dead-simple dictation. Devs hook on the press-to-talk flow, transparent pulsing UI, and input source switching for clean pastes—no voice input not working glitches. Beats github vox box alternatives by injecting into any app, like voice input for claude code or notes.

Who should use this?

Multilingual Mac devs dictating code comments or docs, bilingual writers toggling voice input in whatsapp, or researchers needing voice input devices for classical Chinese output. Ideal for Apple Silicon users frustrated by voice input samsung keyboard habits on Mac, or teams exploring voice github audiobook workflows without Windows 11 voice input limits.

Verdict

Promising early-stage tool (85 stars, 1.0% credibility) with crisp docs and Makefile bundling, but alpha quirks demand OminiX-API setup and accessibility perms—test on M-series Macs before relying daily. Roadmap eyes Windows/Linux; grab it for voice input experiments now.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.