I'm assuming, perhaps incorrectly, that Google would have one backend service or set of services for all speech to text translation, and that all related products would be powered by it. They've been iterating on speech to text interfaces for at least 15 years now.