nine_k 5 hours ago

I suppose you can do it right now. Speech models are excellent.