Google has introduced Gemini 3.5 Live Translate, an innovative AI speech-to-speech model that offers near real-time voice translation in over 70 languages. This new technology distinguishes itself from conventional translation tools by initiating translation while the speaker is still talking, maintaining a delay of just a few seconds. This feature helps retain the speaker's tone, rhythm, and pitch, resulting in more fluid and natural interactions. Additionally, the model is capable of automatically detecting languages and functions effectively even in loud environments. The new feature is now accessible through Google Translate on both Android and iOS devices, with further integration into Google Meet and developer APIs expected to follow soon.
Informational material. 18+.