Skip to main content
WHISPR Desk

Choosing a transcription engine

Applies to WhisprDesk 1.7.3 · Updated 2026-10-25

Choose local transcription and your voice never leaves your Mac. Prefer a cloud model? Bring your own key. Your choice. Both are set up in the same place, Settings > Transcription in the WhisprDesk app, and you can switch at any time.

Before you start

Open the WhisprDesk app: click its icon in the Dock, or open it from your Applications folder. The WhisprDesk window opens with Home / History selected in the sidebar. If WhisprDesk is running but you don't see its window, click the WhisprDesk icon in the menu bar at the top right of your screen and open the window from its menu. Settings is at the bottom of the sidebar on the left.

Transcription models (Whisper, NVIDIA Parakeet, and cloud providers) are set up in Settings > Transcription. AI text cleanup models are set up separately in Settings > Intelligence (see AI cleanup). There is no screen called "model setup".

Local or cloud

Local Cloud
Where your audio goes Stays on your Mac Sent directly from your Mac to the provider you choose
What you need One model download An API key from that provider
Cost Included Billed by the provider to your own account
Works offline Yes, once the model is downloaded No

WhisprDesk does not run its own transcription servers. In cloud mode, your audio goes straight to the provider whose key you entered, under that provider's terms.

The app starts in Cloud mode until you choose a local model. Nothing is transcribed on your Mac until you download one.

Open the transcription settings

Where: WhisprDesk > Settings > Transcription

  1. Open the WhisprDesk app: click its icon in the Dock (or open it from your Applications folder).
  2. Click Settings at the bottom of the sidebar.
  3. Click Transcription.
  4. At the top you will see a banner showing your active model, and below it two tabs: Cloud and Local.

Where: WhisprDesk > Settings > Transcription > Local tab > NVIDIA Parakeet tab

NVIDIA Parakeet is the easy local pick: one download of about 680 MB, it detects the language automatically, and it supports 25 languages. If you rely on the Dictionary for names and special terms, note that Parakeet does not use it; a local Whisper model does.

  1. Open the WhisprDesk app: click its icon in the Dock (or open it from your Applications folder).
  2. Click Settings at the bottom of the sidebar, then click Transcription.
  3. Click the Local tab.
  4. Click the NVIDIA Parakeet tab.
  5. Find Parakeet TDT 0.6B and click Download. The download is about 680 MB and can take a few minutes depending on your internet connection. You can click Cancel while it downloads.
  6. When the download finishes, click the model card. Downloading alone does not select it.
  7. The card now shows In Use, and the banner at the top shows the model with Active. Your next dictation uses it.

Parakeet supports: Bulgarian, Croatian, Czech, Danish, Dutch, English, Estonian, Finnish, French, German, Greek, Hungarian, Italian, Latvian, Lithuanian, Maltese, Polish, Portuguese, Romanian, Russian, Slovak, Slovenian, Spanish, Swedish, and Ukrainian. For other languages, use a Whisper model.

Set up local transcription with Whisper

Where: WhisprDesk > Settings > Transcription > Local tab > OpenAI Whisper tab

  1. Open the WhisprDesk app: click its icon in the Dock (or open it from your Applications folder).
  2. Click Settings at the bottom of the sidebar, then click Transcription.
  3. Click the Local tab.
  4. Click the OpenAI Whisper tab.
  5. Pick a size and click Download. Base is marked Recommended.
  6. When it finishes, click the model card. It shows In Use.
Model Download size
Tiny 75 MB
Base 142 MB
Small 466 MB
Medium 1.5 GB
Large 3 GB
Turbo 1.6 GB

Whisper models download from Hugging Face, and Parakeet from the sherpa-onnx project on GitHub. After that, local transcription needs no internet connection.

Set up cloud transcription

Where: WhisprDesk > Settings > Transcription > Cloud tab

  1. Open the WhisprDesk app: click its icon in the Dock (or open it from your Applications folder).
  2. Click Settings at the bottom of the sidebar, then click Transcription.
  3. Click the Cloud tab.
  4. Click a provider tab: OpenAI, Groq, Mistral, or Custom.
  5. Paste your key into API Key. No key yet? Click Get key. It opens the provider's website, where you sign in and create one.
  6. Click a model. It becomes your active model, and the banner shows Active.

Full details: Cloud API keys.

How selecting a model works

There are no checkboxes in the model picker. Instead:

  • Clicking a model makes it your active model immediately. This is true in either tab. If you have a cloud key set up and you click a downloaded local model, you are now using that local model.
  • Switching between the Cloud and Local tabs also switches your engine. If you click Local and a local model is already selected, your next dictation uses it.
  • Clicking a provider tab (for example Groq) does not change anything by itself. Clicking a model inside it does.
  • A local model must be downloaded before you can select it. A cloud model stays locked until you enter that provider's key.

Check which model is active

Where: WhisprDesk > Settings > Transcription (banner at the top)

  1. Open the WhisprDesk app: click its icon in the Dock (or open it from your Applications folder).
  2. Click Settings at the bottom of the sidebar, then click Transcription.
  3. Look at the banner at the top. It shows the provider, the model name, and Active.
  4. If it shows No model selected instead, dictation will not work until you download and click a local model, or enter a key and click a cloud model.

In the local list, the active model shows In Use. Other downloaded models show Downloaded.

Delete a model

Where: WhisprDesk > Settings > Transcription > Local tab

  1. Open the WhisprDesk app: click its icon in the Dock (or open it from your Applications folder).
  2. Click Settings at the bottom of the sidebar, then click Transcription.
  3. Click the Local tab, then the OpenAI Whisper or NVIDIA Parakeet tab.
  4. Hover over the downloaded model and click the trash icon.
  5. Confirm Delete Model. You can download it again later.

To remove every downloaded model at once:

  1. Open the WhisprDesk app: click its icon in the Dock (or open it from your Applications folder).
  2. Click Settings at the bottom of the sidebar, then click System.
  3. Find Model cache and click Clear Cache. This removes Whisper, Parakeet, and local AI cleanup models.

Set your transcription language

Where: WhisprDesk > Settings > Preferences > Language

  1. Open the WhisprDesk app: click its icon in the Dock (or open it from your Applications folder).
  2. Click Settings at the bottom of the sidebar, then click Preferences.
  3. Under Language, set Transcription language to the language you speak. Auto-detect is the default.

The list has 58 languages. Parakeet supports the 25 listed above. With a cloud provider, supported languages depend on that provider.