Free macOS utility

Turbo Whisper

Private voice dictation for any Mac app. Press OptionSpace, speak, and Turbo Whisper turns your voice into text on your Mac.

  • Tap the shortcut to toggle dictation, or hold it for push-to-talk.
  • See the words appear in a live overlay while you speak.
  • Paste into the app you were using, with a clipboard fallback when automatic paste is unavailable.

Free. Requires macOS 14 or later. A model downloads the first time you select it; speech recognition then runs on your Mac. Want broader workflow context? Read the Mac voice workflows guide.

Dictate without changing how you work

Keep writing in Mail, Messages, Notes, a browser, or your editor. Turbo Whisper handles the voice-to-text step from the menu bar.

Global shortcut

Speak from any app

Use OptionSpace by default. Tap once for hands-free dictation and again to finish, or hold the shortcut for push-to-talk.

Live feedback

Watch your words appear

A compact overlay shows that Turbo Whisper is listening and previews the transcript without taking you away from the app underneath.

On-device

Speech stays local

After the selected model is downloaded, recognition is performed on your Mac instead of sending each dictation to a transcription server.

How to use it

  1. 1

    Install and choose a model

    Open Turbo Whisper from the menu bar and select a speech model. The model downloads once and is cached on your Mac.

  2. 2

    Press Option–Space and speak

    Tap the shortcut for hands-free dictation, or keep it held while you talk for push-to-talk. The shortcut can be changed in Settings.

  3. 3

    Finish the thought

    Tap again to stop a hands-free session, or release after push-to-talk. Turbo Whisper finishes the text on your Mac.

  4. 4

    Carry on typing

    With optional Accessibility permission, the result is pasted into the app you were using and your previous text clipboard contents are restored. Otherwise, the result is copied for you to paste manually.

Need backup before you erase a Mac? Use Mac Backup Guru to create a bootable copy first.

Four on-device speech models

Choose the model that suits your Mac, language, and preference. Availability varies by hardware and macOS version.

Apple Neural Engine

Parakeet and Nemotron

Parakeet TDT 0.6B v3 supports multilingual dictation. Nemotron Streaming 0.6B is an English-only, lower-latency option. Both require Apple silicon and macOS 14 or later.

Apple silicon

Qwen3 ASR 0.6B

A multilingual option built on FluidAudio’s Qwen3 speech-recognition pipeline. It requires an Apple-silicon Mac running macOS 15 or later.

Broad compatibility

Whisper Large V3 Turbo

WhisperKit provides the broadest Mac compatibility in Turbo Whisper and supports multilingual recognition. Try the available models with your own voice and workload.

Frequently asked

Which speech models can I choose?

Turbo Whisper offers Parakeet TDT 0.6B v3, Nemotron Streaming 0.6B, Qwen3 ASR 0.6B, and Whisper Large V3 Turbo. Hardware, operating-system, and language support differ by model.

Does Turbo Whisper work without the internet?

You need an internet connection the first time a selected model is downloaded. Once it is cached, voice recognition runs locally and does not require a routine audio upload.

Which permissions does it need?

Microphone permission is required to hear your dictation. Accessibility permission is optional and enables automatic paste into the app you were using; without it, Turbo Whisper copies the result to the clipboard.

Can I change the keyboard shortcut?

Yes. OptionSpace is the default, and you can record a different global shortcut in Turbo Whisper’s Settings.

Free app, small studio

Turbo Whisper is maintained by the same one-person shop behind Mac Backup Guru and Install Disk Creator. Got feedback or bugs? Please share them.

Contact support Follow on X @BenedictSlaney Enjoying it? Bitcoin tips: 1MbwUFSbQvMrd826DTpup7PpTTNrrJjaJ8