Privacy Mode (local transcription) processes all recording and transcription entirely on your device. Data is not uploaded to the cloud — it is saved only in the local folder you choose.
It is designed for confidential meetings, legal or compliance discussions, and any recording where you need to keep data off external servers.
Prerequisites for using this feature
Plan: Pro, Business, and Enterprise plans (availability may vary by billing method)
Platform: Notta Desktop
Note: If you would like to use Privacy Mode with a paid plan via invoice payment, please contact your sales representative.
How does Privacy Mode Recording work?
Privacy Mode runs as a dedicated local transcription workspace inside Notta Desktop. It is a mode switch that applies to the entire app — you cannot apply it to individual recordings selectively. To record or transcribe in Privacy Mode, turn the mode on before you start.
Before you enter Privacy Mode, you choose a local storage folder and install an offline speech recognition model. After setup is complete, Notta Desktop can process all recordings locally on your device without relying on external servers.
In Privacy Mode:
- Audio transcription is processed locally on your device.
- Recordings and transcripts are saved only in the local folder you choose.
- Local transcription is available once the offline model has been installed.
- Cloud transcription, cloud AI summaries, cross-device sync, and team collaboration are paused or limited.
Recordings and transcripts generated in Privacy Mode are saved to your local folder and are not automatically uploaded to Notta's cloud. Exiting Privacy Mode does not trigger any automatic upload of previously saved local data. If you want to use local data in the cloud, you need to manually export the file and re-import it in cloud mode.
Because Privacy Mode uses an offline speech recognition model installed on your device, recording and transcript content is never sent to Notta's cloud speech recognition service and is not used for AI model training. Edits you make to a transcript apply only to that local record and do not affect the underlying speech recognition model.
Tips:
- Privacy Mode limits how Notta handles transcription data, but it does not block all network activity. You may still need an internet connection for sign-in, eligibility checks, app updates, and offline model downloads.
- The only network dependency during offline use is periodic license validation (roughly once a month). Once validated, you can use Privacy Mode fully offline for at least one month with no outbound traffic during recording or transcription.
- If license validation cannot be completed after the offline period expires, you may need to reconnect to the internet before continuing to use Privacy Mode.
What can I do in Privacy Mode?
You can use Privacy Mode to:
- Start a local real-time recording and transcription.
- Import audio or video files from your computer and transcribe them locally.
- View and manage Privacy Mode records on the same device.
- Export transcripts or files when needed.
Records created in Privacy Mode remain on the device where they were created. They are not automatically synced to other devices or to the Notta cloud workspace.
Files stored in your selected local folder may still be affected by your device settings, backups, file-sharing tools, or organization-managed IT policies.
System requirements
To use Privacy Mode Recording, make sure you have:
- The latest version of Notta Desktop.
- A device that meets the requirements for the local transcription engine you choose.
- Enough free disk space for the offline speech model and local recordings.
- Internet access during initial setup, sign-in, eligibility checks, offline model download, and periodic license validation.
Notta Desktop offers two local transcription engines designed for different device environments and language needs:
| Local transcription engine | Device requirements | Supported languages |
|---|---|---|
| Offline Transcription Model | 1.0 GB storage, 4 GB RAM, and 2 CPU cores | English, Japanese, Simplified Chinese, and Cantonese |
| Apple Speech Analyzer | macOS 26 or later only; 27.3 MB storage, 1 GB RAM, and 2 CPU cores | 11 languages on macOS 26 and later: English, Japanese, Simplified Chinese, Traditional Chinese, Cantonese, Korean, German, French, Spanish, Italian, and Portuguese |
If your device does not meet the recommended requirements for the selected local transcription engine, Notta may warn you before downloading or using that engine.
How to set up Privacy Mode Recording
- Open Notta Desktop.
- Go to Desktop Settings.
- Choose a local storage folder. This is where Privacy Mode recordings, transcripts, and related files will be saved.
- Install an offline speech recognition model.
- Select Enter Privacy Mode.
After you enter Privacy Mode, Notta Desktop switches to the local transcription workspace. A banner will appear to indicate that Privacy Mode is active.
How to start a Privacy Mode recording
- Enter Privacy Mode from Desktop Settings.
- In the Privacy Mode workspace, start a real-time recording or import an audio/video file.
- Keep Notta Desktop open while recording or transcription is in progress.
- Review the transcript after transcription is complete.
- Export the transcript or file if you need to use or share it outside Privacy Mode.
If you exit Privacy Mode, new recordings return to the normal cloud-based workflow. Existing Privacy Mode records remain on your device and can only be viewed in Privacy Mode.
What happens to cloud features in Privacy Mode?
Privacy Mode focuses on local recording and transcription. Some cloud-based features are paused or limited, including:
- Cloud transcription
- AI summaries that require cloud processing
- Sync across devices
- Team sharing and collaboration
To use these cloud features again, exit Privacy Mode and continue recording in the normal Notta Desktop workflow.
Recording consent
Notta Desktop supports Bot-Free Recording in Privacy Mode, so you can record and transcribe meetings without adding a Notta Bot.
Before recording, you are responsible for following applicable recording laws, workplace policies, and meeting consent requirements. Make sure all participants are informed and have given consent when required.
Troubleshooting
Check that:
- You are signed in to Notta Desktop.
- Your plan supports Privacy Mode.
- You have selected a local storage folder.
- An offline speech recognition model has been installed.
- You are using the latest version of Notta Desktop.
Check your internet connection and available disk space, then try again. If the model files are incomplete, delete the model and download it again.
If your license has not been validated recently, reconnect to the internet and open Notta Desktop to complete license validation. After validation, Privacy Mode can be used offline again for at least one month.
Offline transcription performance depends on your device. If your computer has limited memory or CPU resources, transcription may take longer. Close unnecessary apps or try a lighter supported model if one is available.
Privacy Mode records are stored locally and are not synced to the cloud. Open Notta Desktop on the same device and enter Privacy Mode to view them.