Annotension

Install on your Mac

Download the DMG, drag Annotension into Applications, and open the app.

Release status: the public installer is not available yet. The steps below apply when the download is released.

A standard Mac installation

  1. Open the DMG

    Use the published release when it becomes available.

  2. Drag to Applications

    Copy Annotension.app onto the Applications shortcut.

  3. Eject the disk image

    Wait for the app to finish copying.

  4. Open Annotension

    Launch it from Applications and create a project.

The app requires an Apple Silicon Mac and macOS 13 or later. The hardware target is M2 with 16 GB RAM or above; performance depends on the dataset and workload.

Manual annotation, the gallery, local clustering, and detection dataset import/export do not require Python, Xcode, a source checkout, or an account.

Label your first image →

Optional detector setup

Training and prediction need a separate runtime. The small DMG does not bundle Python packages or pretrained model weights.

Requirements and setup

You need Apple Silicon, macOS 14 or later, and Python 3.11. Initial package and model downloads need internet access and additional disk space.

After installing Python 3.11, open Terminal and run:

sh "/Applications/Annotension.app/Contents/Resources/Set Up Detector.command"

If python3.11 is not on your PATH, pass its full executable path as the first argument. The setup installs pinned packages into your account’s Application Support folder, outside the app bundle.

Open Train Detector… and choose Download Model once, about 81 MB. Training and prediction then use local files. No dataset upload is required.

Training loss is not detection accuracy. Validation accuracy is not calculated yet.

Preview builds and macOS checks

A DMG containing PREVIEW BUILD.txt is ad-hoc signed and not notarized. A downloaded preview may be blocked by macOS. Wait for the signed, notarized release for routine installation; do not disable macOS security protections.