Repository navigation
Murmurly 0.1.0
The first public release.
Murmurly is a dictation and meeting recorder for your Mac. Speech is recognized
on your Mac: Murmurly never uploads your recordings or text to be transcribed.
What it does
- Dictation into other apps. Press ⌥Space, speak, press ⌥Space again: the
text is pasted where your cursor is. You can change the shortcut or use a
double press of ⌘ or ⌥ instead. - Meeting recording. Recording starts on its own when another app starts
using the microphone, or when you press ⌥⇧R. Your microphone and the call
audio are saved as two separate tracks. - Transcripts. After a meeting, Murmurly writes a transcript with
timestamps, with your microphone and the call audio labeled separately. - Files. Transcribe audio and video files: mp3, m4a, wav, mp4, mov.
- History. Every recognized dictation is kept in a searchable history.
- Two speech engines. Parakeet (about 0.5 GB, 25 European languages) or
Whisper (about 1.6 GB, about 100 languages). Models download from Hugging
Face only when you ask. - Interface in 12 languages. All translations except English and Russian
were made by machine; corrections are welcome. - Automatic updates. Murmurly checks GitHub for a new version once a day;
you can turn this off in Settings → Permissions. An update you choose while a
meeting is being recorded or transcribed, a file is being transcribed, or
during a dictation, is installed when you quit instead of restarting the app.
Requirements
- A Mac with Apple silicon.
- macOS 15 Sequoia or later.
Before you start
- Automatic meeting recording is on by default. In many countries,
recording a conversation requires the consent of everyone in it, and Murmurly
does not notify other participants. You can turn automatic recording off on
the first screen or in Settings → Meetings. - If macOS won't open the app, this release is not signed with a Developer
ID. Try to open it once, then open System Settings → Privacy & Security and
click Open Anyway. With such a release you may need to grant the
permissions again after each update.
Known limitations
- The other participants' voices are not told apart: everyone on the other side
is "Others". - Dictation and meeting recording never run at the same time. One dictation
lasts at most 5 minutes; an automatic meeting recording at most 4 hours. - A meeting transcription does not resume after quitting. Choosing Quit during
a transcription, or while its transcript is copied to your transcript folder,
warns first; logging out, restarting or shutting down the Mac interrupts it
without a warning. Use Transcribe Meeting… to transcribe it
later. - Removing duplicated speech from the transcript does not work for languages
written without spaces, such as Chinese and Japanese. - ⌥Space is also the default shortcut of ChatGPT for Mac. If you use both,
change one of them.
The full list is in the
README.