Voice to text, entirely on-device.

Name: open-wispr
Author: human37

Hold a key, speak, release. Your words appear at the cursor. Powered by whisper.cpp with Metal acceleration. 100% local — no cloud, no accounts, no telemetry. Free and open source under MIT.

Install on Mac

curl -fsSL https://raw.githubusercontent.com/human37/open-wispr/main/scripts/install.sh | bash

Free & open source forever. Leave a tip →

How it works

Everything you need. Nothing you don't.

Hold

Press and hold the Globe key. Recording starts instantly. A waveform animates in your menu bar.

Speak

Talk naturally. Whisper runs on your GPU via Metal. No network latency, no cloud processing.

Release

Let go. Transcribed text is typed at your cursor. Works in any app. The temp audio file is deleted.

Compare

Why open-wispr?

	open-wispr	VoiceInk	Wispr Flow	Superwhisper	MacWhisper	Apple Dictation
Price	Free forever	$39.99	$15/mo	$8.49/mo	$69 lifetime	Free
Open source	MIT	GPLv3	No	No	No	No
100% on-device	Yes	Yes	No	Yes	Yes	Partial
Works in any app	Yes	Yes	Yes	Yes	Own app only	Yes
Push-to-talk	Yes	Yes	Yes	Yes	No	No
Whisper engine	Yes	Yes	Cloud AI	Yes	Yes	Apple model
Account required	No	No	Yes	Yes	No	Apple ID
AI features	No	AI assistant	AI rewriting	AI formatting	No	No
Languages	99	100+	100+	99	100+	60+
Install method	Homebrew	App Store	App Store	App Store	App Store	Built-in

Install flow

Guided setup. Under a minute.

Paste the install command. It taps the Homebrew formula and installs the binary. Everything downloads behind a spinner.

Grant Accessibility permission. The installer opens System Settings for you and waits. Toggle the switch, come back.

Grant Microphone permission. A system dialog appears. Click Allow. One time only.

Model downloads automatically. The default base.en model (142 MB) downloads on first launch. You can change this later.

Service starts. A waveform icon appears in your menu bar. Hold Globe, speak, release. Done.

Features

Nothing extra. Nothing missing.

Private

Audio is recorded to a temp file, transcribed locally, and deleted. Zero network requests. No accounts, no telemetry, no cloud.

Fast

Runs on Apple Silicon GPU via whisper.cpp and Metal. Transcription is real-time or faster on M1 and above.

Universal

Text appears at your cursor in any app. Editors, browsers, terminals, Slack, email. Missed your focus? Copy Last Dictation from the menu bar.

Configurable

Choose your hotkey, model size, and language. One command to change, one restart to apply. Everything in the terminal.

Multilingual

Supports 99 languages via Whisper's multilingual models. Set an ISO 639-1 code in the config and use a non-.en model.

Lightweight

No Electron. No app bundle to drag. Installs via Homebrew, runs as a background service. Sits quietly in your menu bar.

Configuration

One file. Six options.

~/.config/open-wispr/config.json

{
  "hotkey": { "keyCode": 63, "modifiers": [] },
  "modelSize": "base.en",
  "language": "en",
  "spokenPunctuation": false,
  "maxRecordings": 0
}

Option	Default	Values
hotkey	63	Globe (63), Right Option (61), F5 (96), or any key code
modifiers	[]	`cmd` `ctrl` `shift` `opt` — combine for chords
modelSize	"base.en"	`tiny.en` 75 MB, fastest · `base.en` 142 MB, good balance (default) · `small.en` 466 MB, more accurate · `medium.en` 1.5 GB, great accuracy · `large` 3 GB, best accuracy (M1 Pro+ recommended). Larger = more accurate but slower.
language	"en"	Any ISO 639-1 code. Use a non-.en model for other languages.
spokenPunctuation	false	Say "comma", "period", etc. to insert punctuation manually instead of auto-punctuation.
maxRecordings	0	Optionally store past recordings locally as `.wav` files for re-transcribing from the tray menu. 0 = nothing stored (default). Set 1-100 to keep that many recent recordings.

Edit ~/.config/open-wispr/config.json, then restart: brew services restart open-wispr

What people say

"I use this with Claude Code every day. Hold the key, say what I want, and it just types it out. No more switching between keyboard and some dictation app. It's perfect for long prompts."

Tanner Thorne

Software Engineer

FAQ

Common questions.

Is open-wispr really free?

Yes, completely free and open source under the MIT license. No subscriptions, no usage limits, no premium tier. You can leave a tip if you want to support development.

Does any audio or text leave my computer?

No. Everything runs 100% on-device. Audio goes to a temp file, transcribed locally by Whisper, then deleted immediately. There are zero network requests -- no cloud, no accounts, no telemetry. Optionally, you can configure open-wispr to store a number of past recordings locally via the maxRecordings setting. Those recordings stay private and on your machine, and we default to not storing anything.

How does this compare to VoiceInk, Whisper Flow, or Superwhisper?

open-wispr is the free, MIT-licensed alternative. VoiceInk ($39.99, GPLv3) is the closest — also open source and local — but adds AI features, custom dictionaries, and app-specific modes. Wispr Flow ($15/mo) uses cloud AI. Superwhisper ($8.49/mo) is local but closed-source. open-wispr gives you the same core experience — push-to-talk local dictation powered by Whisper — with zero cost, zero bloat, and the most permissive license.

What Macs are supported?

Any Mac with Apple Silicon (M1 or later) running macOS 13+. Whisper uses Metal GPU acceleration for fast inference. Intel Macs are not supported.

How accurate is it compared to Apple Dictation?

Significantly better. Whisper handles punctuation, technical jargon, code terms, and accented speech far more reliably than Apple's built-in dictation. It supports 99 languages and produces properly punctuated output by default.

Can I use it in any app?

Yes. open-wispr types transcribed text at your cursor position, so it works in any application — editors, browsers, terminals, Slack, email, and more. No plugins or integrations required.