The TTS model needs to be downloaded before audio can be generated.
Go to Settings > TTS to download models.