⏳ In Brief
- Google adds Live Translate for real-time speech with transcripts and audio.
- New Practice mode generates goal-based listening and speaking exercises.
- Live Translate supports 70+ languages, handles accents and noisy places.
- Availability today for Live Translate in the US, India, Mexico.
- Practice beta begins with specific pairs, plus progress tracking and stars.
Translate adds real-time conversations and adaptive practice
Google is rolling out Gemini-powered features to Translate, adding Live Translate for natural back-and-forth speech and a Practice beta that adapts lessons to your skill level and goals. The update targets travel, study, and everyday conversation.
Announced August 26, 2025, Live Translate is available today in the US, India, and Mexico. The Practice beta starts with English↔Spanish or French, and English for Spanish, French, and Portuguese speakers.
Live conversations, now inside the Translate app
Live Translate lets two people speak normally while the app plays translated audio and shows bilingual transcripts. It recognises pauses and intonations, then switches speakers automatically, which keeps chats flowing during travel or daily errands.
Google says advanced voice and speech recognition help isolate sounds in noisy cafés and airports. Support spans 70+ languages, including Arabic, Hindi, Korean, Spanish, French, and Tamil, accessed from the app’s Live translate button.
“We’re bringing two new features to Translate to help with live conversations and language learning.”
Practice mode, tailored lessons that meet you where you are
Practice creates listening and speaking sessions on the fly, tuned to your level and motivations, for example travel, study, or family conversations. You can pick Basic, Intermediate, or Advanced, then the app adapts difficulty as you progress.
Activities include tapping words you hear in short clips, or speaking with an AI partner to complete small tasks. Daily progress is tracked, with stars for completed rounds, plus controls to make exercises easier or harder between rounds.
What users will notice first
- Goal-based scenarios, travel, school, work, or family chats
- Adjustable difficulty, plus a visible daily progress tracker
Availability, languages, and where the beta starts
Live Translate launches first in the US, India, and Mexico, with more regions expected later. It works on Android and iOS, and uses Gemini’s reasoning to improve quality, speed, and multimodal handling.
The Practice beta is available to English speakers learning Spanish or French, and to Spanish, French, or Portuguese speakers learning English. Supported countries include the United States, United Kingdom, Australia, Singapore, Nigeria, and South Africa.
You can change proficiency levels, delete practice history, and submit feedback in-app. These controls matter as generative AI remains experimental, which the help page flags explicitly.
Why this matters for AI language tools
Translate now covers quick lookups, real-time dialogue, and structured practice in one place. That mix reduces app switching, and makes casual learners more likely to sustain speaking habits with immediate, context-aware feedback.
For global users, noise-robust speech and bilingual transcripts improve accessibility in crowded settings. Early beta limits keep expectations clear, which helps trust while Google scales languages and regions responsibly.
Conclusion
Translate’s Gemini update turns a utility into a coach, handling live talk and adaptive lessons with clear controls. The first wave focuses on reliable speech, readable transcripts, and practical scenarios that reflect real-world conversations.
As rollouts expand, watch for additional language pairs, broader country support, and deeper personalisation. If quality holds in noisy places and lessons stay motivating, Translate could become many users’ default learning companion.
For more AI stories, visit AI News on our site.
