BabelVoice Voice Notebook
A privacy-first voice notebook for recording, importing, and transcribing audio with local storage on your device.
Privacy first
Local storage
Privacy first Local models on device
BabelVoice
Record locally, import audio, and turn speech into text in one tap. BabelVoice uses on-device capabilities and local models so recordings, voice audio, and transcripts stay on your device by default and are not uploaded to our servers.
Recordings stay on device by default
WhisperKit local transcription
No user audio uploads
BabelVoice
Voice Notebook
Meeting notes
Follow-up ready
Transcribed
Copy ready
WhisperKit Tiny
Local first
From voice to text, with fewer steps.
BabelVoice is built around one flow: open, capture, and turn audio into text on device. It prioritizes local processing, does not upload user recordings or voice audio, and keeps model downloads, copy actions, and export feedback in context.
Local recording
Record directly on device. Finished clips appear at the top of the list for fast meeting, class, and idea capture.
Audio import
Bring existing audio into BabelVoice, then manage, play, transcribe, and export it from one place.
Offline transcription
Use Apple on-device recognition or WhisperKit local models. First-use downloads show progress on the current recording page, and transcription is designed to stay on device whenever possible.
Local storage
Successful transcripts are stored locally and restored when you reopen the app. They update only when you transcribe again.
Quick copy
Copy the whole transcript from the detail page and get a light confirmation without losing context.
Flexible export
Export TXT, SRT, and VTT for notes, subtitles, archives, and editing workflows.
A complete path for every recording.
The interface keeps users in context: the list stays light, while downloads and transcript results live on the recording detail page.
01
Record or import
Start from the main record button or import existing audio from the top-right action.
02
Choose a model
WhisperKit Tiny is the default. When users switch models, BabelVoice remembers the latest choice.
03
Transcribe
If the model is not ready, the detail page shows download progress before transcription continues.
04
Copy or export
Text stays local, can be copied anytime, and exports into common text and subtitle formats.
Privacy first by design.
BabelVoice is designed for local voice notes. Recordings, imported audio, voice content, and transcripts stay on device by default. Model files download as needed. We do not upload your recordings, voice audio, or transcripts for product analytics.
Privacy first by default
No sale of personal data
No user recording or voice uploads
Microphone permission is for recording
Speech permission is for transcription