I think there is a bug in the speech module.

I am using the Pythonista app on an M1 mac.

On Apple silicon macs (and maybe also on certain models of iPhone/iPad ?), speech recognition is done on-device. (To achieve this, the language model is downloaded and the processing is done on-device, and audio data is not sent to Apple servers)

When using the speech.recognize() command in a Pythonista script, to recognize speech from an audio file, I have noticed that it works correctly only if my computer is not connected to the internet.

If my internet connection is active, then I get the following error:

Traceback (most recent call last):
  File "<string>", line 1, in <module>
RuntimeError: Speech recognition failed: The operation couldn’t be completed. (kAFAssistantErrorDomain error 203.)

I have to turn off my wifi, so that Pythonista does not have internet access, in order for the speech recognition to work.