-
Notifications
You must be signed in to change notification settings - Fork 15
GUI Installation
Requires macOS 15+. Homebrew installs the app, ffmpeg and yt-dlp:
brew install --cask mizorewww/tap/course2md-guiAlternatively, download the DMG, open it and drag course2md.app to Applications. Manual installs also need brew install ffmpeg yt-dlp. Release apps and disk images are Developer ID signed and notarized by Apple.
There is no prebuilt Intel Mac GUI; use the CLI.
For x86_64, install the GUI with its video tools and graphics libraries:
yay -S course2md-gui-bin
# Or: paru -S course2md-gui-binOpen course2md from your applications menu, or run course2md-desktop. Install a Vulkan driver for your GPU; vulkan-swrast provides a software rendering option.
course2md-gui-bin provides the window and private engine; course2md-bin provides the terminal command. They can coexist.
-
Download the GUI ZIP and extract everything to a permanent folder.
-
Install video tools in PowerShell:
winget install --id Gyan.FFmpeg -e winget install --id yt-dlp.yt-dlp -e
-
Restart the app/terminal, then open
course2md-desktop.exe. Keepcourse2md.exebeside it.
Download and extract the GUI tar.gz, enter its folder and run ./course2md-desktop. Keep the bundled course2md beside it.
On Ubuntu 24.04, install:
sudo apt update
sudo apt install ffmpeg yt-dlp libxcb1 libxkbcommon0 libxkbcommon-x11-0 libfontconfig1 libx11-6 libwayland-client0 libvulkan1 mesa-vulkan-driversThe archive is built on Ubuntu 24.04; older distributions may have incompatible glibc versions. Linux ARM64 currently has CLI builds only.
- Choose a notes folder and recognition method. Missing tools lead to installation help.
- Check Settings → Runtime environment. Restart the app after installing tools.
- Add a video URL or local file, check the source and export formats, then generate notes.
- Read transcripts, screenshots and files in the library; organize courses using lists, cards and folder groups.
Available subtitles take priority by default. Apple-native recognition downloads models on first use. Local GPU/CPU recognition needs llama-server and models; cloud APIs need your endpoint, key and model. See recognition and models and cloud and AI.
Bilibili QR login is available in account settings and the video source page. See the Bilibili guide.
GUI and CLI share configuration. Valid GUI settings save automatically. Moving or deleting logical folders does not delete notes. Task progress, cancellation and error details are available in the app.
Homebrew: brew update && brew upgrade --cask course2md-gui. AUR: yay -Syu or paru -Syu.
For manual installs, quit the old app and replace the whole app/archive. Do not mix engine versions. Preserve settings, notes and model caches. Upgrade standalone CLI separately.
Uninstall with brew uninstall --cask course2md-gui or sudo pacman -R course2md-gui-bin. For manual installs, remove the application folder. You do not need to delete personal notes, settings or model caches.
- The GUI bundles its engine but does not add a
course2mdterminal command. Install the CLI separately for scripting. - GitHub's Source code archives are not application installers.
- For missing tools, failed model downloads or graphics startup problems, see troubleshooting.