Google is rolling a new AI-operated experimental feature in Google translation, designed to help people practice and help a new language to learn, the company announced on Tuesday. Translation is also gaining new live abilities to make it easier to communicate with a different language speaking person in real time.
The new language practice feature is designed for both beginners who are starting to learn convenient skills and want to brush advanced speakers on their vocabulary, the company says. To do this, it creates listening and speaking practice sessions that suit the user’s skill level and unique learning goals.
With this new language practice feature, Google is carrying on Duoolingo, the popular language learning app uses a gameified approach to help users to practice in more than 40 languages.

To use the feature, you will choose the “Practice” option in the Google translation app. From there, you can set skill levels and goals. Google translation then produces customized landscape where you can either hear conversations and tap the words you understand, or you can practice speaking. Practice track users’ daily progress, says Google.
Beta Experience is rolling out in the Google translation app for Android and iOS starting on Tuesday. This feature is first available for English speakers to practice Spanish and French, as well as practicing English for Spanish, French and Portuguese speakers.
The Google translation is also introducing the ability to users to conversate back-end-world with audio and on-screen translations through the translation app.
Google wrote in a blog post, “Building our existing live conversation experience, our advanced AI models are now making it easy to make live conversations in more than 70 languages - including Arabic, French, Hindi, Korean, Spanish and Tamil.”
Techcrunch event
San francisco
,
27-29 October, 2025

You can tap the “Live Translate” option in the translation app and then choose the language you want to translate only by speaking. Then you will listen loudly with a translation of your interaction in both languages. The app will translate and switch between the two languages you and the other person are speaking.
Google notes that this feature can identify stagnation, accents and intonations to allow for natural-sounding conversation.
This feature uses Google’s voice and speech recognition model to separate sounds, meaning that you will be able to use live capabilities in a loud restaurant or busy airport.
These live translation capabilities begin on Tuesday for users in the US, India and Mexico.
Google wrote in its blog post, “These updates are possible with progress in AI and machine learning.” “As we carry forward the limitations of language processing and understanding, we are able to serve a wide range of languages and improve the quality and speed of translations. And with our Gemini models in translation, we are able to make heavy progress in translation quality, multimodal translation and text-to-speakes (TTS) capabilities.”
Google says that people translate about 1 trillion words into translation, search, lens and circle. The company believes that these new AI-operated facilities will help in removing language obstacles.