I downloaded the VisionDemo.zip again, but still no sound.. Should it show the text overlay of the spoken audio like in the screenshots? It's not doing that either. It only shows "Scanning the environment" for a few seconds and then nothing happens.
Yes it should. I've just tested on another PC and the current build works for me.
Should work like in this demo
But if there is no text it sounds like a problem with the OpenAI server connection.
There should also be an audio file received from OpenAI:
There are two more demos you can test to make sure your API key is valid: