Streaming Microphone Input to Google Speech API

I looked into the Google Cloud Speech API and got a streaming microphone running on a Node.

I was then wondering what would be the best practice for streaming my microphone from the web interface? Does it post the audio stream from getUserMedia to the Node server and pass it to the API with the Node API client? Or is it just saving the voice input to a file, which I then pass to the API?

The goal is to "decode" the instructions (one or two sentences long) and send the result to another API.