Check who can hear translated audio
DeepL said on November 11 that it had begun using Aure to generate translated speech in DeepL Voice. For an online meeting, the practical limit is at the listener’s end. DeepL’s help center says each participant who wants spoken translation needs the desktop app and an eligible Voice license. A team can use that requirement to work out who would be able to hear translated audio before it invites outside participants.
The company says Aure adjusts the delivery of the translation as the speaker changes pace or stresses particular words. It lists English, French, German, Italian, Japanese, Korean, Portuguese and Spanish for the initial rollout. DeepL says it applies the model automatically for supported languages in online meetings, in-person conversations and its Voice API. The company continues to use earlier models for other languages.
Test the phrases that carry decisions
For a pilot, CGH would use a short meeting script with dates, names, quantities and a few sentences where emphasis changes the meaning. Ask bilingual listeners to check whether the translated audio preserves both the instruction and the speaker’s emphasis. A more natural voice is a separate measure from whether people heard the right deadline or amount. CGH reviewed the documentation and did not test the model.
DeepL says online spoken translation is unsupported in mainland China and unavailable through its web app. Its in-person product page directs buyers to sales for a quote. The launch materials reviewed gave no separate Aure price or full country-by-country rollout list.