Google is launching a significant update for its Gemini audio models, enhancing the Google Translate app with live speech-to-speech translation capabilities. This upgrade features the Gemini 2.5 Flash Native Audio model, optimized for complex voice interactions. The beta experience, currently available, allows users to hear real-time translations through headphones, making it ideal for travel and overcoming language barriers.
The new functionality offers continuous listening for group conversations and two-way dialogues, allowing seamless language exchange. The standout feature, “style transfer,” mimics the speaker’s tone and speed, ensuring natural-sounding translations. With support for over 70 languages and 2,000 language pairs, the app automatically detects and translates multiple languages simultaneously.
Additionally, the revamped audio model enhances performance and conversation fluidity, making it an essential tool across Google products like Google AI Studio and Search Live. The beta version is currently accessible on Android in the US, Mexico, and India, with broader support expected soon.
Source link
