Google Meet’s speech translation is finally coming to your phone. After rolling it out on the web earlier this year, Google is now extending near real-time speech translation to the Google Meet apps on Android and iOS, so you can join multilingual calls on the go without worrying about language barriers.

The feature does more than just slap text captions at the bottom of your screen. Google’s AI actually dubs the speaker’s voice into another language, keeping their tone and cadence as much as possible, while lowering the original audio in the background. That makes the conversation feel closer to having a human interpreter in the room, instead of reading along like a subtitle track.

On mobile, it works much like on desktop: you join a meeting, choose your translation language pair, and Meet will start translating spoken English into one of the supported languages (or vice versa). Right now, Google supports bidirectional translation between English and five major languages: Spanish, French, German, Portuguese, and Italian. So you can have, say, a US-based speaker talking in English while a teammate in Brazil hears everything in Portuguese, directly through their phone’s Meet app.

There are a few limitations worth knowing. At the moment, only one language pair can be active per meeting, so you can’t run a single call with multiple parallel interpretation tracks. It’s more like “this meeting is English ↔ Spanish” or “English ↔ German,” not a full-blown multilingual conference setup yet. Also, users joining from meeting room hardware can listen to translations, but their own speech won’t be translated, which is something hybrid teams will want to keep in mind.

Under the hood, Google is using a speech‑to‑speech translation model that transcribes what you say, translates it, and then synthesizes a new voice in roughly two to three seconds—fast enough that the conversation still feels natural. This is built directly into Meet, so there’s no separate bot joining the call or any extra plugin to install.

On the rollout side, Google is doing its usual staged launch. For Rapid Release Google Workspace domains, the feature is rolling out gradually starting April 8, 2026, and can take up to 15 days to appear. For Scheduled Release domains, the rollout starts April 23, 2026, again with up to a 15‑day window. So if you don’t see it today on your phone, it may simply not have hit your domain yet.

Availability is fairly broad, but not universal. On the business side, speech translation is coming to Business Standard, Business Plus, Enterprise Standard, Enterprise Plus, and Frontline Plus. On the consumer side, it’s available for people subscribed to Google AI Pro and Google AI Ultra, and it’s also included with add-ons like AI Ultra Access and Google AI Pro for Education. In other words, this is positioned as a premium AI feature, not a freebie for every basic Meet user.

From a real-world angle, this could be a big deal for remote and hybrid teams that work across regions. Instead of hiring interpreters or juggling third‑party tools, a single person with an eligible plan can enable translations so everyone on the call benefits—desktop or mobile. For education and training, it makes it much easier to run global workshops or guest lectures where the speaker sticks to their native language, but students still follow along in theirs.

If you’re an admin, speech translation is on by default and can be toggled at the organizational unit level in the Admin console, so you can roll it out to specific teams first (for example, sales or support teams that work across markets). End users can then head to the Meet settings to pick their language and learn the exact steps from Google’s Help Center guide for Speech Translation in Meet.

For now, Google says users can expect ongoing improvements—both in the UI and in the quality and nuance of translations—as the system learns from more use and as the underlying AI models evolve. It’s also clear from Google’s research that they’re already looking at expanding beyond the initial set of Latin‑based languages to more complex language pairs in the future.