Root and Runtime
SoftwareRoot and Runtime
SoftwareVersion 0.2.0 released

Speakerdesk

Meeting transcripts with speaker labels, processed on your Mac

Download for Apple Silicon

Version 0.2.0 · macOS 14.0+ · 98 MB

Release notes

Real-meeting capture and packaged transcription testing are still pending.

Verify the download

SHA-256

b03fb37f3478583400916de5dfaed7a8557d90b34c6479d79821a663fa2339c9
  • Meeting transcripts
  • Local processing
  • macOS
Availability
Available now
Runs on
Apple Silicon · macOS 14.0 or newer
Setup
About 2.3 GB for model setup, plus space for recordings
Audio and transcripts
Stored and processed on your Mac
Speakerdesk in light and dark themes. Names and transcript text are sample content.

Overview

What it is

Speakerdesk turns a meeting or an audio file into a transcript you can review. Follow the conversation by speaker, correct the text, and listen back to the recording.

Audio processing runs on your Mac. The first setup downloads the transcription models; there is no paid transcription service to connect.

Capabilities

What it does

  1. 01

    Follow a live meeting

    Capture microphone and system audio, with text appearing in short passages as the conversation progresses.

  2. 02

    Work from a recording

    Import an audio file and review the transcript alongside playback.

  3. 03

    Keep track of speakers

    Separate the conversation into labeled speakers, then apply names and correct passages before exporting.

  4. 04

    Reuse saved names

    Save people locally and choose their names in future meetings. Naming a person does not save their voice or identify them automatically.

  5. 05

    Read in light or dark

    Use a light or dark interface for the transcript, playback, and speaker controls.

Using Speakerdesk

Good to know

  1. Privacy

    Keep audio processing local

    The app uses local transcription and speaker-labeling models. Internet access is needed to download the models during setup; recordings and transcripts stay on the Mac.

  2. Names

    Let you decide who is speaking

    Saved names and speaker corrections work today. Automatic recognition of a person's voice is unavailable in this release.

  3. Review

    Make the transcript your own

    Review the text and speaker labels before sharing a transcript. You can correct passages, change names, and listen back to check what was said.

Choose a saved name or name a speaker in the current meeting. Sample name shown.

Release notes

Version 0.2.0

Available for Apple Silicon Macs running macOS 14.0 or newer. The Mac app is signed with Developer ID and notarized by Apple.

Real-meeting capture and transcription from the packaged app have not yet been tested. Review transcripts and speaker labels before sharing them.

Saved names are chosen by you. Automatic recognition of a person's voice is unavailable. The first setup downloads the local transcription models.

Now readingWhat it isSectionsClose
On this page