How do I select input and target languages?

The input language is the language being spoken. It defaults to automatic detection, or you can search and select it in the input-language list. The target language is the language you want to read or hear and must be chosen separately. These lists serve different purposes: the target-language list does not establish every supported input language. Use the corresponding lists in the current realtime workbench.

On a computer, each capture round has one target language. On a phone, each turn is translated into the one language selected for the other person; the same speech is not output into multiple target languages at once. Changes to input language, target language or voice apply to the next round and do not retranslate completed sentences. To make multiple language versions of an existing file, use file translation.