Google expands AI-powered language tools, including real-time translation and transcription

Expanding Language Support

Google's technologies and products now support interactions in over 300 languages, covering more than 7 billion people globally. This expansion addresses the historical underrepresentation of many languages in the digital world, a goal Google has pursued since launching Google Translate in 2006.

Advancements in AI for Understanding

Google is moving beyond text-based translation to native audio intelligence, training models like Gemini to process audio directly. This approach aims to capture elements of human communication such as tone, pacing, emotion, and context, which are often lost in traditional multi-step speech recognition systems.

This direct audio processing allows for better handling of natural speech patterns, including laughter, hesitations, and code-switching between multiple languages within a single sentence.

New Real-time Dialogue Tools

Gemini 3.5 Live Translate now provides real-time spoken translation across 70 languages and over 2,000 language pairs. This tool is designed to naturally capture code-switching and emotional cues during conversations.

Gemini 3.5 Transcribe is Google's most precise speech-to-text model to date. It converts raw audio into formatted text, functioning effectively in noisy environments and with complex jargon. This technology also powers features like Rambler on Android Gboard, which refines spoken input by removing filler words, correcting grammar, and allowing voice-based editing and language switching.