Google has started rolling out a new voice simultaneous translation tool for Google Meet applications on Android and iOS operating systems. The technology uses artificial intelligence to process participants’ speech in real time during video conferences. The feature has been operating in the version for web browsers since last year. Agora functionality reaches mobile devices to facilitate remote communication anywhere.
The system is powered by the Gemini model, which delivers translated audio almost instantly to listeners. Essa update aims to break down language barriers in global teams and educational environments. The tool goes beyond traditional text subtitles. The processed audio tries to maintain the original characteristics of the speaker, generating a more natural and fluid conversation experience directly via the cell phone or tablet.
Como artificial intelligence preserves the tone of voice on calls
The main innovation of the feature lies in the ability to clone vocals during live transmission. The software does not deliver a generic robotic voice. Artificial intelligence analyzes the user’s speech pattern and generates a synthetic version in the target language. The original sound remains at a very low volume in the background, while the dubbing takes the lead in the call.
Essa’s technological approach ensures that the emotional context of the conversation is not lost in translation. Profissionais who use the platform are able to transmit their ideas with the same original intensity. The tool captures subtleties of human communication efficiently.
- The system identifies and reproduces moments of excitement or seriousness during speech.
- The synthetic voice adapts the rhythm and emphasis of the words according to the chosen language.
- Apenas A member of the meeting must have the compatible plan to release the role to others.
The operating dynamics only require the participant to activate the option and define the desired language for listening. Processing takes place in the Google cloud. Translated audio is delivered with minimal delay, allowing debates and presentations to follow a continuous flow without sudden interruptions that impair understanding.
Release Prazos and packages required for operation
The implementation of the new feature in mobile applications takes place gradually throughout the month of April 2026. Corporate Domínios configured for rapid release began receiving the update on April 8, 2026. Já environments with scheduled release will have access to the tool from April 23, 2026. Users need to keep the application updated in the official stores of their operating systems.
Access to voice translation requires specific subscriptions to the Google Workspace ecosystem. The company restricted the new feature to the most robust packages aimed at corporate or advanced educational use. Conventional free Contas do not have the ability to initiate the vocal cloning feature.
Eligible plans include the Business Standard, Business Plus, Enterprise Standard, and Enterprise Plus. Assinantes from the Google AI packages Pro and Ultra also receive the functionality. In the education sector, the tool is part of the Education Plus plan. If the organizer or a participant has one of these licenses, everyone present in the virtual room will be able to listen to the translations at no additional cost.
Idiomas available in initial phase and system expansion
Neste firstly, technical support covers bidirectional pairs between English and five other languages of great global circulation. The official list includes Spanish, French, German, Italian and Portuguese. Essa initial selection covers a significant portion of the commercial and academic interactions carried out daily on the video platform.
The company has already confirmed that the language base will grow in the coming months. The Turk figures to be the next confirmed addition to the development schedule. Engenheiros software works on refining artificial intelligence models. The goal is to increase accuracy in detecting regional accents and complex idiomatic expressions.
Expanding the language catalog depends on continuous training of the Gemini model. The quality of simultaneous translation improves as the system processes new linguistic data. Google remains focused on delivering dubbing that respects the grammatical rules and cultural context of each region supported by the application.
Configurações Privacy and Activation on Mobile Devices
Voice data security represents a priority in the architecture of the new Google Meet feature. Translation activation does not occur automatically or hiddenly. The user needs to grant explicit permission through the specific button on the application interface. Qualquer participant has the freedom to revoke this authorization at any time during the video call.
Para To enable the function on Android or iOS, the process requires just a few taps on the smartphone screen. The user must access the menu represented by three dots during the video conference. Then navigate to the settings area and locate the voice translation activation key. The intuitive interface makes it easy to use for people with different levels of technical knowledge.
Administradores of corporate networks have full control over the availability of the tool. The management panel allows you to block or release the resource for specific groups of employees. Google applies strict security protocols to ensure that processed audio is not intercepted or improperly stored during simultaneous translation sessions.
