Google has introduced Gemini 3.5 Live Translate, an AI model that enables near-instant voice-to-voice translation during multilingual conversations. The tool supports more than 70 languages and aims to reduce delays common in traditional systems. It became available to developers on Tuesday.
The model performs continuous streaming translation rather than processing speech in turns. This approach allows conversations to proceed with only a few seconds of delay while preserving the original speaker's pacing, intonation, and emotional tone. Google said the system handles noisy environments, overlapping voices, and informal speech. It automatically detects languages and supports thousands of pairings within a single conversation. Developers can access the model through a public preview in the Gemini Live API and AI Studio. Select enterprise customers will receive it in Google Meet this month, with a wider release planned later. The tool will also reach the Google Translate app on Android and iOS devices soon. All audio streams include SynthID watermarks to indicate they are AI-generated. The company emphasized that the technology is designed for practical use in settings such as customer support, tours, and classrooms.
