Google Meet is the odd one out among the big three. Zoom and Teams both stop at translated captions for most people, and Zoom’s voice translation still can’t do German. Meet will translate your actual speech, out loud, in something approximating your own voice, and it’s been doing it for a while.
It also has the widest caption language list of the three by a distance: 100+ target languages, where Teams manages around 30.
So why isn’t this the obvious answer for events? Because of a 90-minute timer, five language pairs, and one line in the docs about livestreams. Details below, all linked to Google’s own documentation. Prices are from Google’s European pricing page in EUR, correct as of writing. Google prices regionally, so check your own.
What you get without paying anything extra
Captions. Meet captions the meeting in the language being spoken, on every edition including free accounts. Google expanded the spoken-language list beyond English to French, German, Portuguese and Spanish and has kept adding since.
That’s the whole free tier. Everything else here is paid.
Which Workspace edition unlocks translation
Google’s premium Meet features page is the authoritative list, and it splits three ways:
- Translated captions: Business Standard, Business Plus, Enterprise Standard, Enterprise Plus, Education Plus, or a Google One AI Pro / AI Ultra subscription.
- Transcripts: the same editions, plus Google One Premium (2 TB and up).
- Speech Translation: Business Standard, Business Plus, Enterprise Standard, Enterprise Plus, Frontline Plus, AI Pro for Education, and Google One AI Pro / AI Ultra.
The thing to notice: Business Starter gets none of it. Translation starts at Business Standard.
Google’s European list prices, per user per month, billed monthly:
| Edition | Price (EUR) |
|---|---|
| Business Starter | €6.80, no translation |
| Business Standard | €13.60, translated captions + Speech Translation |
| Business Plus | €21.10 |
| Enterprise | quote only |
Note the Business tiers cap at 300 users; above that you’re on Enterprise and talking to sales, same as Zoom.
A worked example. Fifty people who need translated captions, moving from Business Starter to Business Standard: the difference is €6.80 per user per month, so €340/month or €4,080/year on top of what you already pay. Unlike Teams, there’s no organiser-only shortcut. The entitlement is per user, and you can’t buy it for five people and have it cover everyone else’s meetings.
That comparison is worth sitting with. Teams charges $10/user/month for translation but only for the organiser. Meet charges €6.80/user/month more, for everybody. For a small team of organisers Teams is dramatically cheaper; for an organisation where everyone runs meetings, Meet is competitive.
How to turn on translated captions in Google Meet
In the meeting: More options → Settings → Captions. Set Language of the meeting to what’s actually being spoken, turn on Translated captions, then pick your target language.
The “Language of the meeting” field is the one that matters. Same trap as Teams: set it wrong and you get fluent nonsense rather than an error.
Google’s documentation for translated captions puts the coverage at over 100 languages, including all the majors and a long tail well past what Zoom or Teams offer. You can also restyle the captions (font, size, colour) and scroll back through what’s already been said, which neither competitor does.
Speech Translation: the one that talks
This is Meet’s genuinely distinctive feature, and Google’s Speech Translation documentation describes it as translating your speech in real time “in a voice like yours”.
To start it: Meeting tools → Speech translation, choose the language combination, then Enable translation for everyone. Other participants then confirm their own language and grant permission for their voice to be translated. Consent is built into the flow, which is more than you can say for most of this category.
Now the limits, and they’re substantial:
- Five language pairs. English paired with French, German, Italian, Portuguese or Spanish. That’s it.
- One language pair per meeting. Not one per participant, one per meeting. A genuinely multilingual room is out.
- A 90-minute limit. Google states it plainly. Most conference sessions fit; a conference day does not.
- Not in livestreams or recordings. If your event is streamed, Speech Translation doesn’t apply to it.
- A few seconds of delay, which Google says is deliberate. It waits for complete phrases rather than translating word by word. That’s the right trade for accuracy and the wrong one for interruption-heavy conversation.
- Meet hardware is receive-only. People in a room on a Meet device can hear translations but can’t have their own speech translated.
- Not for under-18s.
Compared to Zoom’s voice translation, Meet is ahead where it counts for Europe: German works today, where Zoom’s is still promised for later in 2026. Compared to Teams, there’s no contest. Teams has no machine voice translation at all, only human interpreters.
Where the limits actually are
The three features have three different language lists. Captions cover one set of spoken languages, translated captions cover 100+ targets, and Speech Translation covers five pairs. “Google Meet supports 100+ languages” is true of exactly one of the three.
Business Starter is a dead end. Cheapest paid Workspace tier, zero translation. Anyone budgeting from the Starter price will be wrong by a factor of two.
No organiser-only licensing. The Teams trick, licensing five organisers to cover everyone, has no equivalent here.
Livestreams are excluded from Speech Translation. This is the one that rules Meet out for a lot of events, because streaming is how large events reach people who aren’t in the room.
Captions live in the client. Same structural limit as Zoom and Teams: captions are for the person on their own laptop. There’s no room display, no second screen, nothing an audience can read collectively.
And it’s still machine translation. Cross-talk, accents, jargon and product names degrade it; two error rates stack when you translate a transcription. Google’s own framing, a few seconds behind for completeness, is an honest admission that real-time and accurate pull in opposite directions.
Would I rely on it?
For an internal bilingual team meeting where everyone’s on a laptop and the session is under 90 minutes, Meet is the strongest of the three platforms and it isn’t especially close. Speech Translation is a real feature, not a demo, and 100+ caption languages is genuinely useful.
For an event, four questions kill it quickly:
- Longer than 90 minutes? Speech Translation stops.
- More than two languages? One pair per meeting.
- Streamed? Speech Translation doesn’t apply.
- Is anyone in a room rather than on a laptop? Meet hardware can listen but not be translated, and there’s no shared display.
Which is the same conclusion as the other two platforms, reached by a different route. In-meeting translation is built for people already in the call on their own device. Making an event accessible means a caption feed the room can see, a transcript people follow on their own phone in their own language, and a record you can point at afterwards. Our use cases section covers what that looks like, and the accessibility statement sets out the standards we hold ourselves to.