What happened
Google released Gemini 3.5 Live Translate, an audio model delivering near real-time speech-to-speech translation across over 70 languages. This model continuously generates fluid, natural-sounding translated speech, preserving intonation and pacing, unlike turn-by-turn systems. It handles multilingual inputs and noise robustly, staying just seconds behind the speaker. Gemini 3.5 Live Translate is rolling out via public preview for developers through the Gemini Live API and Google AI Studio, and in private preview for enterprises in Google Meet. It is also available in the Google Translate app on Android and iOS, with all generated audio watermarked by SynthID.
Why it matters
Real-time, natural language barriers reduce across global operations. Platform engineers gain API access to integrate continuous, low-latency translation into applications, supporting over 70 languages and 2000+ language combinations in tools like Google Meet. This mechanism reduces communication friction for international teams and customer interactions. Procurement teams evaluating communication platforms must now factor in this enhanced capability, which follows OpenAI's recent voice model advancements. Security architects should note the SynthID watermarking for AI-generated audio, providing a verifiable origin for translated content.
