According to Jam Jam Online, this system automatically recognizes languages and produces speech that reflects the tone, speed and pitch of the real speaker’s voice.
Unlike traditional turn-based translations, Gemini 3.5 Live Translate produces audio continuously with a delay of only a few seconds to ensure semantic accuracy and preservation of speech context, while remaining in sync with the speaker’s words.
According to Gadget, this technology is being offered in various Google products. Enterprise users will also have access to the private preview version of this feature in Google Meet from this month.
For normal users, this feature will be available on Android and iOS operating systems. This integration supports more than 70 languages, which is a significant expansion compared to the previous limitations.
Android users will benefit from a new feature called “Listening Mode” in the Google Translate application.
This feature allows users to hear translations directly through their phone’s speakerphone; A feature that is useful for private listening without the need for headphones. Google Meet will also be upgraded and will support more than 2000 language combinations in a single meeting; A huge leap from the service’s previous capabilities, which were mainly focused on the English language. The purpose of these changes is to provide smooth and uninterrupted communication at the global level.
The resistance of this model to environmental noise and its ability to manage multilingual inputs without the need for manual settings are among its key features for use in real situations. These capabilities make it suitable for simultaneous interpreting in meetings, training classes and live broadcasts.
Initial feedback from business partners shows that the translation quality, accuracy and low latency of this technology have been very impressive.
All AI-generated sounds will be digitally tagged with SynthID technology to be identifiable and prevent the spread of false information.
