Google Meet Debuts AI-Powered Real-Time Voice Translation

At Google I/O 2025, Google announced a real-time speech translation feature for Google Meet, powered by Gemini AI and DeepMind's language audio model. The tool overlays the speaker's original voice at low volume with AI-generated translations, aiming to replicate a simultaneous interpreter and preserve the speaker's identity. Initial language support includes English and Spanish, with Italian, German, and Portuguese to follow soon. The beta rollout targets AI Pro and Ultra subscribers, with Workspace enterprise customers to receive early access later in the year. The launch has sparked interest in AI-related cryptocurrencies and tech stocks and highlights Google's broader AI ambitions, such as advanced text-to-speech and personalized smart replies. Early demos showcased both the technology's strengths and challenges in handling regional dialects and slang.