Google's Gemini 3.5 Live Translate drops the Pixel Buds requirement Google's newest AI model lets anyone hold a real-time conversation across languages using nothing more than a smartphone, no special hardware required. Gemini 3.5 Live Translate processes speech continuously as someone talks, translating and speaking back in the other person's language with only a few seconds of delay.
Previous real-time translation efforts from Google demanded specific hardware like Pixel Buds or a Pixel phone. Ars Technica reported that last year's Translate app update still required the company's earbuds on Android.
Gemini 3.5 Live Translate works with any headphones, and Android users can skip earbuds entirely by holding the phone to their ear like a normal call using a new Listening Mode. The model, part of the 3.5 family that debuted at Google I/O, supports 70-plus languages and automatically detects which language a person is speaking.
Google Product Manager Anuda Weerasinghe and Senior Staff Software Engineer Tony Lu wrote in the announcement that the system balances the trade-off between waiting for context to improve quality and translating immediately to stay in sync with the speaker. The result preserves intonation, pacing, and pitch rather than producing robotic output.
Google is rolling the feature out across four products simultaneously. The Google Translate app on Android and iOS gets Live Translate with support for any connected headphones.
Google Meet expands from 5 languages to over 70, supporting more than 2,000 language pairings in a single meeting, though enterprise customers get access as a private preview this month with wider availability planned for the latter half of 2026. Developers can start building with a public preview in the Gemini Live API and Google AI Studio.
Constellation Research's Holger Mueller told SiliconANGLE the model may deliver better quality than some human translators. "The release of Gemini 3.5 Live Translate shows it has not yet relinquished that lead, both in terms of translation quality and supported languages," he said.
"Now it's pushing the envelope further again with simultaneous translation in a consumer app for the first time, and it may even be better quality than some human translators."
Grab, the ride-hailing giant, is already piloting the technology for communication between drivers and passengers. The company's Chief Product Officer Philip Kandal cited automatic multilingual detection and low-latency voice translation as key advantages for Grab's 10 million-plus monthly voice calls.
All audio generated by Gemini 3.5 Live Translate carries SynthID watermarks embedded directly in the waveform data, making AI-generated speech detectable and irremovable. Google says the model handles background noise and overlapping voices in busy environments. A Pro version of the 3.5 translation model is expected in the coming weeks.
