Importing audio (advanced)

Importing existing audio and watching Voice Memos aren’t in the app yet — they run through the optional command-line engine. Set it up once:

git clone https://github.com/MeetGist/meetgist.git
cd meetgist
make setup                          # Python env for the engine
# then put your key in scripts/.env:  GEMINI_API_KEY=your-key

A file or folder

Transcribe audio you already have (.m4a / .mp3 / .wav / .mp4):

python3 scripts/transcribe_file.py ~/Downloads/interview.mp3

The documents are written in place, next to the source file (interview.transcript.md, interview.polished.md, interview.summary.md). Pass a folder to transcribe every audio file inside it. A one-voice memo stays single-speaker; a conversation becomes Speaker 1 / Speaker 2 / … .

Apple Voice Memos (auto-import)

Watch your Voice Memos library and auto-transcribe new memos from your iPhone, Apple Watch, or Mac:

scripts/install_voicememo_watcher.sh        # start (checks every ~10 min)
scripts/install_voicememo_watcher.sh status
scripts/install_voicememo_watcher.sh uninstall

On install it marks existing memos as seen, so only memos recorded from then on are imported. The launchd agent’s Python interpreter needs Full Disk Access to read the Voice Memos store.