Translate a Google Meet call in real time

Google Meet has invested heavily in live captions. The captions are good, in English, and decent in a handful of widely-spoken languages — and they’re useful for following along while watching the speaker. They are not a replacement for listening.

Reading captions while a meeting moves quickly is cognitively demanding. The listener stays one or two sentences behind, misses the connecting beats, and stops trying to participate. The same listener, given translated audio in their ear, follows in real time. Loquira sits beside Google Meet to provide that audio without changing anything about the Meet call.

How the setup works

Google Meet’s audio is closed in the same way Zoom’s is. The participants share a single mix. Loquira does not try to enter that mix. Instead, the host runs a parallel Loquira session on a second device — typically a phone within arm’s reach — that hears the host’s voice through its own microphone. Listeners join Loquira on their phones or laptops and select their language.

The setup is identical in shape to the Zoom setup. The differences are administrative: Meet links live in Google Calendar, chat history disappears when the call ends, and Google Workspace tiers gate certain features but not the host’s ability to run a second app on a second device.

Recommended setup

• Microphone. Whatever you use for Meet. A USB condenser or a wired headset gives Loquira the cleanest signal. The audio requirements page explains the signal quality the recognition pipeline expects.
• Loquira device. A phone on your desk, within a metre of your mouth. If you’re presenting in a conference room with multiple speakers around a table, position the phone where it can hear all of them — or assign each speaker their own Loquira session and let listeners switch.
• Where to put the link. The Calendar event description is the primary place. Meet’s chat is a fallback. Attendees who join late won’t see chat messages from before their arrival, which is why Calendar matters.
• Earbuds for listeners. Same as with Zoom — listening to Loquira through a phone speaker while the Meet audio plays in the background creates echo. Recommend earbuds in the invite.

Multi-speaker calls

Google Meet’s active-speaker detection is reliable, which makes multi-speaker calls easier to translate than they sound. As long as speakers pause briefly between turns and the room is reasonably quiet, the Loquira device picks up each voice as it speaks. For distributed teams where each speaker is on their own tile, the per-speaker session pattern applies — see the Zoom setup for the details, which transfer directly.

What about Meet’s own translation?

Google Meet ships translated captions for a small set of language pairs (en ↔ es, en ↔ fr, en ↔ de, en ↔ pt) on certain Workspace tiers. Where it’s available and the listener prefers captions over audio, it’s a solid option. Where it isn’t available — most language pairs, free Workspace accounts, mobile Meet clients — Loquira fills the gap.

The two tools coexist without conflict: Meet renders its captions on the Meet UI; Loquira renders its audio on the listener’s phone or laptop. A listener can use both, or either, or neither.

Limitations specific to Google Meet

• Phone-dial-in attendees. Attendees who joined Meet by phone cannot use the Loquira listener on the same line. Direct them to the Loquira link from a second device.
• Recording. Meet’s recording captures Meet’s audio only. Loquira’s transcripts and audio replays are separate artifacts; link both in the meeting follow-up.
• Live streaming a Meet to YouTube. If you’re streaming a Meet call to YouTube Live for broader audiences, treat the YouTube layer separately — see the YouTube Live setup.
• Education accounts. Workspace for Education has additional moderation controls that may restrict chat. Use the Calendar invite for the Loquira link; do not rely on Meet chat in classroom contexts.