Download JesType
Free for 48 hours, then €14.95 once. Speech recognition runs on your computer.
How to install
- 1
Download and open the DMG
Use the download button above to get the installer.
- 2
Drag to Applications
Drag JesType into the Applications folder in the DMG window.
- 3
Launch and grant permissions
Open JesType from Applications. Grant microphone access when prompted.
What's included
Offline speech recognition
The Parakeet and Moonshine models run on your device. Transcribing needs no internet connection.
Dictation in your apps
Dictate into the focused text field in email, documents, Slack, your browser or a code editor.
Hold-to-record shortcut
Hold a keyboard shortcut you choose to record. Let go to transcribe and paste.
Word replacements and autocorrect
Add your own replacements and autocorrect rules for technical terms, names or phrases you use a lot.
System requirements
macOS
- macOS 13.4 (Ventura) or later
- Apple Silicon (M1, M2, M3, M4)
- Disk space for the installer and one speech model
- Microphone access permission
- Accessibility permission (for auto-paste)
Windows
- Windows 10 or later (64-bit)
- Disk space for the installer and one speech model
- Microphone access
- No special permissions required
During setup JesType downloads one speech model, about 456 MB for Parakeet V3, about 2.5 GB for Parakeet V3 High quality or about 202 MB for Moonshine Medium Streaming. Leave some extra free space while it unpacks. After that, speech recognition works without a connection. Activation, license checks and updates go online. A paid license keeps working offline for up to 30 days between checks; the 48-hour trial needs a connection.
Speed and memory use depend on the model, the length of the recording and your computer. The free trial is the best way to check yours.
Don't have a license yet? Get one for €14.95 →
Not sure yet? Start the free 48-hour trial in the app with your email address.
Your privacy matters. JesType recognizes speech on your device with the Parakeet and Moonshine models. Recordings are never sent to any server. If you turn on cloud AI cleanup, the transcript text, never the audio, goes to the provider you choose.