
Tech • IA • Crypto
Google has launched Gemini 3.5 Live Translate, enabling real-time speech translation across 70 languages without relying on English as an intermediary.
Gemini 3.5 Live Translate introduces direct translation between 70 languages, eliminating the need for pivot languages like English. The system operates in real time, allowing fluid, natural conversations between speakers of different languages.
The model reproduces intonation, rhythm, and vocal characteristics, making translated speech sound closer to the original speaker. This approach goes beyond traditional text-to-speech, aiming for more natural and context-aware communication.
The technology is being embedded across Google products, including Google Translate on mobile devices. On Android, users can hold their phone to their ear and hear live translations as if in a phone call, enabling discreet, seamless conversations.
In Google Meet, currently for enterprise users, participants can speak in their native languages while hearing others in their own language. This enables multilingual meetings where each user experiences a fully localized conversation in real time.
Early use cases include transport services, where drivers and passengers can communicate instantly across language barriers, and international business interactions that previously required interpreters or intermediaries.
The technology is expected to disrupt sectors such as dubbing, translation, and interpretation, raising concerns about job displacement while also creating new opportunities in AI-assisted workflows.
Google’s rollout builds on similar real-time translation efforts from competitors like OpenAI, positioning Gemini as a faster and potentially more cost-effective alternative in the growing AI translation market.
Real-time, voice-preserving translation is rapidly becoming mainstream, reshaping global communication while introducing both opportunities and disruptions across industries.