For the default layout, Wiktionary doesn't provide that data, so that won't be happening. I'll put in a request and maybe they'll add it one day.

I tried fetching an audio recording of the pronunciation, which worked, but Wiktionary doesn't do a good enough job at classifying the language of the recordings, so sometimes I ended up with Dutch pronunciations for the English word, for example. It was too confusing and misleading, so we won't be seeing that feature yet.

Once Wiktionary improve the API, then that's a different story.