I Stayed Up All Night and Built a Free-for-Life Voice Input Tool

Last night, I just couldn't take it anymore.

Voice input tools on the market today — they all want you to log in, or pay for a membership; they limit usage count, or they cap duration.

Some start out saying they're free, then pop up paywalls once you're hooked.

The worst part: all you want is to convert your voice to text, but you get blocked by every kind of third-party service.

So I just stayed up all night and built one myself.

The name is simple: VoiceInput.

It's a voice input tool for macOS.

You place your cursor in any input field, press a hotkey to start speaking, press it again to stop — it automatically transcribes what you said into text and pastes it into the active window.

The default hotkey is:

Option + K

The whole flow is dead simple:

Place your cursor where you want to type
Press Option + K to start recording
When done, press Option + K again
Speech is auto-recognized as text
Auto-copied to clipboard, and pasted into the current input field if permission allows

This isn't a demo. The full pipeline is already working:

Always-on menu bar icon
Global hotkey to start recording
Microphone audio capture
Real-time speech recognition
Floating panel at the bottom showing status
Recognized text written to clipboard
Auto-paste into the current app when permission is granted
Manual Cmd + V fallback when auto-paste permission isn't available

That means you can use it directly in WeChat, Feishu, browsers, code editors, and documents.

You speak. It types.

That's exactly what I wanted.

Why Did I Build It Myself?

The reason is simple: I refuse to be held hostage by third parties any longer.

Voice input should be a basic capability.

Every day, when you write messages, write documents, write articles, or jot down inspiration — often it's not that you can't write, it's that you're too lazy to type.

Especially for long text, typing is just too slow.

But many voice input tools have overcomplicated this: