Streaming API

Provides access to Hume models through secure web socket connections

You can build real-time applications by running continuous inference on streams of audio, video, and text data.

The Streaming API supports the following Hume models:

  • burst
  • face
  • facemesh
  • language
  • prosody

Navigate to our Streaming API reference page to learn more.


Learn more about the models and about the science


To maintain a connection longer than 1 minute you will need to implement reconnect logic in your application.