SpeechRecognition: audiostart event

The audiostart event of the Web Speech API is fired when the user agent has started to capture audio for speech recognition.


Use the event name in methods like addEventListener(), or set an event handler property.

addEventListener("audiostart", (event) => {});

onaudiostart = (event) => {};

Event type

A generic Event with no added properties.


You can use the audiostart event in an addEventListener method:

const recognition = new webkitSpeechRecognition() || new SpeechRecognition();

recognition.addEventListener("audiostart", () => {
  console.log("Audio capturing started");

Or use the onaudiostart event handler property:

recognition.onaudiostart = () => {
  console.log("Audio capturing started");


Web Speech API
# eventdef-speechrecognition-audiostart
Web Speech API
# dom-speechrecognition-onaudiostart

Browser compatibility

BCD tables only load in the browser

See also