Hold a key, speak, and keep writing. VoxKey is a small, native Mac app that turns speech into text using local Whisper models. No account, subscription, or saved recordings. Includes custom vocabulary and English or multilingual models, including Spanish. Free and open source. Requires Apple Silicon and macOS 26+. Download a model once, then dictate offline.
I'm Rod. I built VoxKey because I wanted a small and native Mac app for offline dictation. I tried OpenWhispr but wanted something simpler that focused just on dictation.
Hold a key, speak, and release to insert the text where you were typing. It's written in Swift and uses WhisperKit to run speech recognition locally. No recordings or dictation history are saved.
It's free and open source (MIT). Requires Apple Silicon and macOS 26+. The app is signed and notarized. You only need to download a speech model during setup and after that dictation works offline. It supports English and other languages, including Spanish, depending on the model.
If you try it, I'd love to hear how it works for you and what gets in the way. You can hit me at hello@rodrigouroz.com or leave an issue in the repo.
Hi,
I'm Rod. I built VoxKey because I wanted a small and native Mac app for offline dictation. I tried OpenWhispr but wanted something simpler that focused just on dictation.
Hold a key, speak, and release to insert the text where you were typing. It's written in Swift and uses WhisperKit to run speech recognition locally. No recordings or dictation history are saved.
It's free and open source (MIT). Requires Apple Silicon and macOS 26+. The app is signed and notarized. You only need to download a speech model during setup and after that dictation works offline. It supports English and other languages, including Spanish, depending on the model.
If you try it, I'd love to hear how it works for you and what gets in the way. You can hit me at hello@rodrigouroz.com or leave an issue in the repo.
Source: https://github.com/rodrigouroz/V...