{"id":265437,"date":"2026-07-03T15:24:38","date_gmt":"2026-07-03T07:24:38","guid":{"rendered":"https:\/\/www.grab.com\/sg\/?post_type=editorial&p=265437"},"modified":"2026-07-03T15:24:38","modified_gmt":"2026-07-03T07:24:38","slug":"live-voice-translations-solving-language-barriers-with-seamless-calls","status":"publish","type":"editorial","link":"https:\/\/www.grab.com\/sg\/inside-grab\/stories\/live-voice-translations-solving-language-barriers-with-seamless-calls\/","title":{"rendered":"Live Voice Translations: Solving language barriers with seamless calls"},"content":{"rendered":"\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t
\n\t\t\t
\n\t\t\t\t\t\t\t<\/div>\n\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/section>\n\t\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t
\n\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t

Imagine a traveller landing in Bangkok who opens the Grab app and books a ride. The driver arrives two minutes early and calls to confirm the pickup point, but neither side speaks the other\u2019s language. The call is confusing, adding friction to what should have been a seamless pickup.\u00a0<\/p>

Southeast Asia is a linguistically diverse region, making situations like this a common occurence. Across our markets\u00a0 travellers and driver-partners make more than 10 million voice calls every month, and many of those are between people who do not share a common language.<\/p>

We saw an opportunity to make those interactions more seamless.<\/p>

The live voice translation challenge<\/h5>

In recent years, text translation has become fast, accurate and widely accessible. But live voice translation is a very different challenge, where low latency and natural-sounding speech are just as important as translation accuracy.<\/p>

At Grab, we\u2019ve tested live voice translations with different models. Typical challenges we encounter are high latency, limited coverage for Southeast Asian languages, and translations that often sound flat or unnatural.<\/p>

Testing Gemini 3.5 Live Translate<\/h5>

Our latest trials have been with Gemini 3.5 Live Translate, a speech-to-speech translation feature that can be integrated into in-app voice calls. The model is promising, as it provides low-latency, two-way translation in over 70 languages and is better at preserving speakers\u2019 natural rhythm, pitch, and intonation. This is necessary for drivers and passengers\u00a0 to understand each other in real-life situations.<\/p>

The feature also auto-detects multiple languages while ensuring low latency. That\u2019s helpful in our region with its diverse heritage, where everyday conversation can involve a mix of languages, such as Malay, English, and Chinese.<\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t

\n\t\t\t\t
\n\t\t\t\t\t
\n\t\t\t
<\/div>\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t
Making progress on seamless voice translations<\/h5>

After several iterative optimisations, the feature integration is starting to work seamlessly for Grab\u2019s use case. We\u2019re observing improvements along key live voice translations engineering challenges:<\/p>

  • For example, we\u2019ve reduced the lag at the start of a voice live translation by \u201cwarming up\u201d all necessary systems before the conversation starts\u2014such as allocating memory and preloading code.\u00a0<\/li>
  • We built a digital filter that ignores silence and background noise, so the system only focuses on actual speech.<\/li>
  • We created a smart playback manager that cuts out old or delayed audio and automatically speeds up or slows down the playback, making sure the translated speech keeps pace with the live conversation.<\/li>
  • We also got rid of turn-based conversations, so there\u2019s no need to wait for the other party to finish speaking before responding. This allows interruptions and lets the conversation flow more naturally.\u00a0<\/li>
  • We added live text transcriptions alongside voice translations on the call screen, giving users the flexibility to read along rather than relying solely on audio.\u00a0<\/li><\/ul>

    We continue exploring solutions that enable better communications between our users and partners across languages and aim to ship a first iteration of the live voice translation feature to Grab users by the second half of this year. With the latest technologies, a future where talking to a driver in Bangkok who only speaks Thai feels just as natural as speaking your own language may not be far away.<\/p>\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/section>\n\t\t\t\t

    \n\t\t\t\t\t\t
    \n\t\t\t\t\t
    \n\t\t\t
    \n\t\t\t\t\t\t\t<\/div>\n\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/section>\n\t\t\t\t<\/div>\n\t\t","protected":false},"parent":180237,"menu_order":0,"template":"grab21-default","acf":[],"_links":{"self":[{"href":"https:\/\/www.grab.com\/sg\/wp-json\/wp\/v2\/editorial\/265437"}],"collection":[{"href":"https:\/\/www.grab.com\/sg\/wp-json\/wp\/v2\/editorial"}],"about":[{"href":"https:\/\/www.grab.com\/sg\/wp-json\/wp\/v2\/types\/editorial"}],"version-history":[{"count":33,"href":"https:\/\/www.grab.com\/sg\/wp-json\/wp\/v2\/editorial\/265437\/revisions"}],"predecessor-version":[{"id":265482,"href":"https:\/\/www.grab.com\/sg\/wp-json\/wp\/v2\/editorial\/265437\/revisions\/265482"}],"up":[{"embeddable":true,"href":"https:\/\/www.grab.com\/sg\/wp-json\/wp\/v2\/editorial\/180237"}],"wp:attachment":[{"href":"https:\/\/www.grab.com\/sg\/wp-json\/wp\/v2\/media?parent=265437"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}