Using AI features

If you are using an Apple silicon-based Mac, you will see an option to enable AI in the Settings > Advanced panel. Turning this on will prompt you to download AI models to your system. VDO uses offline AI models so your data stays private and no internet connection is required after the initial download.

These models are not included with the application by default and are entirely optional. However, enabling AI can significantly improve video search and metadata generation.

Download the model that best suits your needs. More details are available in the System Requirements section of this documentation.

Extracting metadata from videos with AI

Once AI is enabled and the models are downloaded, you can use it to extract metadata such as descriptions and generated titles for your videos. You can do this from the video details section of the collection screen, or as a batch process from Settings > Advanced > AI Tools.

Click the Start button to begin. Processing is done one video at a time and may take 1–5 minutes per video, depending on the video length, available RAM, processing power, and the model you selected. Advanced models generally take longer than basic models, but they typically produce higher-quality descriptions.

If processing takes too long, consider switching to a lower-tier model. If you have thousands of videos, it is recommended that you close other applications and let the process run uninterrupted. You can stop the process at any time and restart it later for the remaining videos. Note that VDO may take a short while to finish the currently running task before it fully stops.