Features

Dictation and hotkeys#

• System-wide dictation

  Start recording from any app and insert the result directly into the active text field.

• Flexible hotkeys

  Use Hybrid, Toggle, or Hold mode; additional shortcuts open history, the workflow palette, or the last result.

• Ordered language hints

  Restrict automatic detection to several languages and order the hints by expected priority.

• Recording overlays

  See recording state, duration, the active workflow, and live partial results when the engine supports them.

• History

  Search previous transcriptions, copy text again, and review the engine, model, and duration.

Statistics#

Statistics summarize your saved dictations: days with entries, saved entries, recorded apps and models, words, and recorded minutes. A chart shows activity per day, followed by the apps and models you use most.

Choose a week, a month, all history, or a custom range. The numbers are computed locally from your history, so deleted entries and entries removed by history retention no longer count.

Dictionary and snippets#

• Dictionary

  Pass specialist terms to supported engines as hints and correct recurring recognition mistakes.

• Snippets

  Expand spoken or typed triggers into prepared text and dynamic placeholders.

• Learn from Corrections · Premium

  Turn clear manual edits made after insertion into new dictionary corrections. Up to Windows 1.0.9, the app calls this feature "Automatic correction learning".

• Sync Dictionary & Snippets · Premium

  Keep dictionary entries and snippets in sync across your devices through a shared cloud folder, for example in iCloud Drive, OneDrive, or Dropbox. Up to Windows 1.0.9, the app calls this feature "Cloud Folder Sync".

Files and recorder#

File transcription processes audio and video files as a queue. The recorder captures microphone and system audio for longer sessions and can transcribe the result afterward.

Models and acceleration#

Transcription engines and models are installed as extensions. Local engines work offline; cloud engines send audio to the selected provider. Supported local models can load automatically, on the CPU, or, depending on the model, with NVIDIA CUDA, Vulkan, or AMD ROCm. Changing acceleration can require a restart.

The currently available models, providers, and actions can change independently of the app version.

Plugins and workflows#

Under Settings > Discover plugins, you add local and cloud engines, AI providers, action plugins, memory, and utilities. Every installed plugin gets its own page in the Settings sidebar, which opens right after the installation. The page shows the plugin's status and settings, such as the API key or models. There you can disable or uninstall the plugin, and update it when a new version is available. Uninstalling keeps the API key, preferences, and downloaded models.

Plugins update separately from the app. By default, TypeWhisper checks once a day, downloads new versions in the background, and uses them after the next restart. You can turn this off under Settings > Account & about > Update plugins automatically.

Workflows connect these building blocks to apps, websites, hotkeys, or a global fallback.

Current add-on catalogConfigure workflows