Gemini 3.5 Live Translate Adds Real-Time Voice Translation to Google AI Studio

Google Gemini 3.5 Live Translate brings near real-time, natural speech translation to Google AI Studio, Google Translate, and Google Meet. This update focuses on reducing the latency between speech and translated output while maintaining the original speaker's tone and prosody. The integration allows developers to build applications that handle multi-language audio streams with significantly less friction than previous iterations of Google's translation models.

What's new

Google Gemini 3.5 Live Translate introduces a specialized architecture designed for low-latency audio processing. As of late 2024, the model supports fluid voice translation that mimics natural human speech patterns rather than the staccato delivery common in older text-to-speech systems. The feature is currently rolling out across Google AI Studio for developers and is being integrated into Google Meet for enterprise users.

Key technical improvements in Gemini 3.5 Live Translate include better handling of interruptions and background noise, which previously caused translation errors in live environments. The system now processes audio in smaller chunks to ensure the translated voice follows the original speaker almost instantaneously. This version also expands the library of supported languages, aiming for parity across Google's primary communication tools.

How it fits your workflow

For filmmakers and content creators, Google Gemini 3.5 Live Translate serves as a high-fidelity tool for rough-cut localization and international collaboration. Producers can use the Google Meet integration to conduct interviews or production meetings with international crews, receiving natural-sounding translations in real-time. This reduces the need for human interpreters during the early stages of pre-production or remote scouting.

In terms of competition, Gemini 3.5 Live Translate competes directly with ElevenLabs Dubbing Studio and the live translation features found in Skype or Zoom. While ElevenLabs focuses heavily on high-end performance for finished content, Google's implementation is designed for utility and speed within a live production environment. It functions as a more integrated alternative to standalone translation apps, particularly for teams already using the Google Workspace ecosystem for project management.

Developers using Google AI Studio can now implement these voice cloning and translation capabilities into custom apps. This allows for the creation of bespoke tools that can translate video scripts or voiceover scratch tracks on the fly, helping editors check the timing and flow of localized content before committing to a final professional dub.

What it costs / how to try it

Google Gemini 3.5 Live Translate is available to developers through Google AI Studio and is rolling out to Google Workspace subscribers using Google Meet and Translate. Access levels depend on your current Google Cloud or Workspace tier.