The kind read-only property of the TextTrack interface returns the kind of text track this object represents. This decides how the track will be handled by a user agent.

A string. One of:

"subtitles"

The cues are overlaid on the video. Positioning of the cues is controlled using the properties of an object that inherits from TextTrackCue, for example VTTCue.

"captions"

The cues are overlaid on the video. Positioning of the cues is controlled using the properties of an object that inherits from TextTrackCue, for example VTTCue.

"descriptions"

The cues are made available in a non-visual fashion.

"chapters"

The user agent will make available a mechanism to navigate by selecting a cue.

"metadata"

Additional data related to the media data, which could be used for interactive views.

In the following example the value of kind is printed to the console.

let video = document.querySelector("video");
let track = video.addTextTrack("captions", "Captions", "en");
track.mode = "showing";
console.log(track.kind);