Google Duo’s captions for audio and video messages are rolling out

We aren’t at the point yet where every Android phone is able to transcribe voice calls in real time, but what Google is doing with its video chat service Duo is a good first step: the app is now enabling captions for recorded voice and video messages.

Our tipster Nick spotted the option on his device, but it appears to be controlled by a server-side switch as it’s not available to us. When enabling captions, a pop-up tells you the audio part of the Duo message will be sent to Google for transcription, but that it won’t be stored beyond that. The transcription also uses your phone’s default language, so it should work in more languages than English. Here’s the corresponding support page.

We earlier reported that software leaks blogger Jane Manchun Wong uncovered a toggle for “Captions for messages” back in late February which would give Google permission to collect users’ audio content for post-process transcription. It looks like the other shoe has dropped.

To be clear, this is different from the Live Captions feature Google has made available to Pixel phones where transcription appears for practically any spoken word on the device (not to mention similar real-time caption features for devices from other OEMs). And it’s not Live Transcribe which is able to take dictation from speakers not on the device. A company spokesperson also clarified that captions are available on both Android and iOS versions of Duo.

It’s a good measure for accessibility and we’d generally like to see more of this, every tech company in the world. Thanks.