Google Meet’s Real-Time AI Translation Feature Revolutionizes Multilingual Conversations

Google has begun rolling out real-time speech translation in Google Meet to mobile devices, marking a significant step in breaking down language barriers during video calls. The feature, which translates spoken words into another language in near real-time while preserving the speaker’s voice, is now available for users on Android and iOS platforms. This expansion follows its initial launch on desktop and extends the tool’s accessibility to a broader audience.

How the Technology Works

The speech translation feature leverages Google’s AI models, developed in collaboration between the Google Meet, DeepMind, and Research teams. According to Google’s official documentation, the system translates speech between English and four other languages: French, German, Italian, and Portuguese. The translation occurs with a slight delay to ensure accuracy, and the output mimics the speaker’s original voice, creating a more natural conversational experience.

Fredric Lindstrom, who leads the audio engineering team for Google Meet, emphasized the feature’s potential to eliminate language as a barrier to communication. In a blog post published on September 11, 2025, Lindstrom shared an example of how the tool has already helped families connect across languages. “The idea is to make sure language isn’t a barrier to communication, whether you’re planning a vacation and talking to someone in another country or chatting with friends or family who don’t speak the same language,” he wrote.

Availability and Limitations

The feature is currently in beta and is not available to all users. Google has stated that translation quality and availability may change during this phase, with improvements expected over time. The tool is accessible to subscribers of Google AI Pro, Google Workspace Business Standard and Plus, Google Workspace Enterprise Standard and Plus, Google Workspace Frontline Plus, AI Pro for Education, AI Expanded Access, and AI Ultra Access.

However, there are notable limitations. Speech translation is not available in live streams or recordings, and participants joining via Meet hardware can only listen to translations—they cannot have their own speech translated. The feature is disabled by default for administrators, though they can enable it for users within their organization. Google has also clarified that no audio is saved during translation, and the models are not trained on users’ voices, addressing potential privacy concerns.

Competitive Landscape

Google’s rollout comes as real-time translation technology gains traction across the tech industry. Competitors like DeepL have also made strides in this space, with DeepL Voice Translation offering similar capabilities. However, Google’s integration into Meet—a widely used platform for business and personal communication—positions it as a direct tool for everyday users rather than a standalone application.

The feature’s expansion to mobile devices is particularly significant, as it aligns with the growing trend of remote work and global collaboration. With more users relying on smartphones for video calls, the ability to translate conversations in real time could reshape how multinational teams, international families, and travelers communicate.

Future Developments

While the current version of speech translation supports only four languages beyond English, Google has indicated that the feature will continue to evolve. The company’s support documentation notes that translation quality and language availability are expected to improve over time. Users are encouraged to check back for updates, as the tool may expand to include additional languages and refine its accuracy.

For now, the rollout to mobile devices represents a critical milestone in making real-time translation a mainstream tool. As Google refines the technology, it could set a new standard for multilingual communication in both professional and personal settings.