Skip to content

Releases

v0.1.2

Choose a tag to compare

@BrunoV21 BrunoV21 released this 03 Nov 09:37

✨ Improvements

🗂️ Sessions

  • Added session deletion with a confirmation dialog. The session is removed from the list once deletion succeeds, and a failure to refresh the list is reported separately.
  • The confirmation dialog supports keyboard navigation and keeps focus inside the dialog while deletion runs.

🖥️ Interface

  • Side panels (processing, notes and transcript) can now be collapsed and expanded.
  • Panel collapse state is remembered between launches.

🛠️ Fixes

  • Reduced memory use when cleaning up audio by processing recordings as a stream instead of loading them fully into memory.
  • Improved echo cleanup between microphone and system audio.
  • Kept microphone and system audio aligned across pause and resume.

💻 Compatibility

  • Supports Windows 10 and 11, x64.
  • Preserves existing recordings, transcripts, settings, credentials, and installed model locations during upgrades.

🛸 Co-authored with Nova

v0.1.1

Choose a tag to compare

@BrunoV21 BrunoV21 released this 30 Oct 10:23

✨ Improvements

🔄 Updates

  • Added in-app update support, with checks at startup and in Settings. Downloading and installing updates requires user action, and installation is deferred while recording or unfinished work is active.

🎙️ Recording and Review

  • Added recording timers.
  • Improved meeting review controls, queue and dropdown feedback, and the notes status display.

🛠️ Fixes

  • Fixed an issue where quitting the application could leave the processing queue paused on the next launch.
  • Constrained automatic language detection during transcription.

💻 Compatibility

  • Supports Windows 10 and 11, x64.
  • Preserves existing recordings, transcripts, settings, credentials, and installed model locations during upgrades.
  • Older installations without in-app updates require a manual installation of an updater-enabled release first.

🛸 Co-authored with Nova

v0.1.0

Choose a tag to compare

@BrunoV21 BrunoV21 released this 27 Oct 14:21

Meet Twix, Compass’s private meeting companion, turning conversations into locally processed transcripts with speaker identification and easy export.

✨ What's New

🎙️ Record Your Meetings

  • Capture microphone and system audio using your selected devices.
  • Stop a recording from the app or the system tray. Recordings and timing information are saved locally.
  • Use the configurable consent reminder before recording. You remain responsible for obtaining consent from participants.

🗣️ Transcribe and Identify Speakers

  • Turn recordings into speaker-labelled transcripts with local transcription and speaker identification.
  • Review transcript text and edit speaker names.
  • Cancel processing jobs and resume them after restarting the app.

📤 Keep and Share Your Notes

  • Export transcripts as JSON, Markdown, text, or SRT.
  • Delete a session from the app to remove its local audio and derived data.

🔐 Local Processing with Compass

  • Download and verify a compatible processing model using a Compass API key. No personal Hugging Face account or token is required.
  • Once a model is installed and verified, process recordings offline. There is no cloud transcription fallback.
  • Recordings and transcripts stay local; Compass receives authentication and model-download metadata.

⚙️ Getting Started

  • Twix supports Windows 10 and 11 on x64 computers.
  • In Settings, save a Compass API key and download a model. Wait for Offline ready before processing a recording.
  • See Setup and diagnostics for model, hardware, local-data, and installer details.

🛸 Co-authored with Nova