Google Translate has introduced a new pronunciation practice feature that uses artificial intelligence to analyze users’ speech and provide immediate feedback on how to improve pronunciation. The feature is designed to let learners practice and refine their pronunciation before speaking with real people in the target language.
How it works
The tool relies on speech recognition and analysis to evaluate spoken input and offer suggestions for clearer or more accurate pronunciation. The functionality is similar to pronunciation exercises offered by Duolingo, but is implemented within the Google Translate app.
Availability and supported languages
According to Google, the pronunciation practice will be available initially in the United States and India, and will support three languages: English, Spanish and Hindi. For now, the feature is built into the Android version of the Google Translate app; it has not been added to the iOS app.
Why it matters
Google says pronunciation practice was one of the most-requested features among users, and the addition is part of a series of updates marking Google Translate’s 20th anniversary. The company has not announced plans or a timeline to expand the feature to more regions, platforms or languages.
Sources and notes
The Verge highlighted that this launch is tied to the 20th anniversary events; Google’s communications indicate the rollout remains limited in scope at present.


