Until today, Google's speech recognition capabilities offered support for 89 languages and locales. Starting today, 30 new ones have been added, with a focus on the emerging markets in Africa and India. However, this brings the total to 119 languages only for a limited number of services at this time, since the Google Translate app and others should be updated in the future.
The updated speech recognition capabilities will be present in Gboard for Android and Voice Search, as well as in the Cloud Speech API that is used by certain third-party voice and video apps. The list of new languages includes no less than eight new Indian dialects, as well as two of Africa's most popular languages, namely Swahili and Amharic. To develop support for these new languages, Google combined human voice samples and its machine learning tech.
An improvement comes to US English speakers as well, who can now use voice dictation to use emojis, for example saying "winky face emoji" instead of selecting the emoji manually.