You can translate meetings in real time in 2026 using three approaches: your platform's built-in translation (Zoom AI Companion, Google Meet captions, or Microsoft Teams Premium), a Chrome extension that overlays captions on an existing call, or a browser-based app like MirrorCaption that works across any browser-based meeting — Zoom, Teams, Meet, or Webex — without a bot joining. This guide walks through every method with step-by-step setup, honest tradeoffs, and a comparison table so you can pick the right approach in under two minutes.
Last quarter, a sales team in Munich ran a 45-minute product demo with a prospective client in Tokyo. Their Zoom plan didn't include the AI Companion translation add-on, so no translation ran during the call. Near the end, the client said "少し時間をください" — a standard polite deferral. The team read it as genuine interest and spent three weeks in follow-up. What felt like enthusiasm was the Japanese business equivalent of "we'll pass." A live translation would have changed that conversation in the room.
Real-time translation isn't a convenience feature. It's a decision-making feature — the difference between reading the room and finding out what was said after the call ends.
- Zoom, Google Meet, and Teams all offer built-in translated captions on eligible plans or add-ons — but each is locked to its own platform.
- Browser extensions add captions to existing meetings in Chrome with no changes to the call, but require IT approval in managed environments.
- MirrorCaption works across any browser-based meeting (Zoom, Teams, Meet, Webex) with 50+ selectable languages, no bot, and no extension to install.
- Sub-500ms streaming means the translation arrives while the speaker is still talking — not after the sentence ends.
- For face-to-face conversations, MirrorCaption's Talk mode runs on mobile Chrome — no video call required.
What "Real-Time" Actually Means for Meeting Translation
"Real-time" covers a wide latency spectrum. There are three meaningfully different speed tiers, and they produce very different in-meeting experiences.
- Word-by-word streaming (under 500ms) — Translation lands while the speaker is still mid-sentence. You can interrupt, ask a clarifying question, or change direction in the same breath of conversation. This is what MirrorCaption delivers via streaming speech-to-text over a WebSocket connection.
- Sentence-by-sentence (2–5 seconds) — A full sentence finishes processing before translation appears. You get the gist in time to respond, but you're always one beat behind. Most platform built-in tools fall here.
- Post-meeting transcription — Tools like Otter.ai and Fireflies generate polished, searchable transcripts after the call ends. Excellent for review and notes; not useful for live decisions.
For most in-call use cases — sales calls, negotiations, multilingual standups — you need tier 1 or tier 2 at minimum. To understand the full difference between these approaches, see our explainer on live captions vs transcripts.
The tier matters most when nuance is high. When a Japanese counterpart says "少し考えさせてください," you have the remainder of the meeting to change tack. A word-by-word tool gives you that. A post-meeting transcript gives you something to follow up on tomorrow — after the deal has already moved.
Option 1 — Use Your Meeting Platform's Built-In Translation
The fastest path to real-time translation is the one already built into your meeting tool. Here's what each major platform offers and where it stops.
Zoom Meeting Translation
Zoom offers real-time translated captions through AI Companion on eligible Zoom Workplace Enterprise plans and as a Translated Captions add-on for other paid Workplace plans. Once the account owner or admin enables automated captions and translated captions, participants can select their viewing language in a live meeting.
- Sign in to your Zoom account at zoom.us and go to Settings → Meeting → In Meeting (Advanced).
- Enable Automated captions (and Full transcript if available on your plan).
- In a live meeting, click Show Captions in the bottom toolbar.
- Click the caption settings icon and select your viewing language.
Where Zoom translation falls short: It only runs inside Zoom. If a client sends a Google Meet link, or your company uses Teams for internal calls, you need a separate solution. Availability also depends on the host account, admin settings, and Zoom's current supported-language list, so verify the exact source and target languages before a high-stakes call.
Google Meet Live Translation
Google Meet provides live captions to all users and, on certain Google Workspace tiers, real-time translated captions. Basic captions (English by default) are available with any Google account. Translated captions require a paid Workspace plan; availability varies by tier.
- Start or join a Google Meet call.
- Click the three-dot menu (More options) at the bottom of the screen.
- Select Turn on captions.
- On eligible Workspace plans, click the CC icon → Settings to change the caption language or enable translated captions.
Where Meet translation falls short: Platform-locked — works only in Google Meet. No transcript export on free tiers. Translation language pairs are limited compared to dedicated tools, and the feature set varies significantly depending on your Workspace plan.
Microsoft Teams Live Translation
Teams Premium includes live translated captions and transcripts — a significant upgrade over standard Teams captions. Teams Premium is listed at approximately $10 per user per month as an add-on to a base Microsoft 365 or Teams license, per Microsoft's current pricing. Verify your organization's rate, as volume licensing can vary.
- During a Teams meeting, click the three-dot menu and select Turn on live captions.
- With a Teams Premium license, click the CC icon and then Change spoken language to set the meeting's source language.
- Each participant can then select their own caption display language from the same CC menu.
Where Teams translation falls short: Microsoft ecosystem only. No benefit for Zoom calls, Google Meet, or in-person conversations. Teams Premium is a per-seat add-on cost that escalates with team size. For a detailed cost comparison, see MirrorCaption vs Teams Premium.
When built-in tools are enough
Platform-native translation is the right call when your entire team is on one platform, the host controls the meeting settings, and you need one or two language pairs that platform supports. When your meetings cross platforms, when 50+ language pairs matter, or when you can't control the host's plan settings, you need a different approach.
Option 2 — Use a Browser Extension
Several Chrome extensions — Tactiq, Notta Sidebar, and similar tools — overlay translated captions on top of your existing meeting tab. They capture audio from your browser and feed it to a speech-to-text and translation pipeline, displaying results in a sidebar or overlay without joining the call.
What works well: Extensions often have a free tier, they don't modify the meeting itself, and other participants see no indication that you're using one. They require no change to host settings.
What doesn't:
- Chrome-only. No Microsoft Edge, no Firefox, no mobile.
- Extension installation requires IT approval in most managed corporate environments — the same environments where you're most likely to need translation.
- Translation is usually an add-on to the core transcription feature, not the primary design goal. Accuracy for non-European languages varies.
- Extension policies change. An extension approved today can be blocked at your next IT review cycle.
Extensions are a reasonable middle ground if you work primarily in Chrome on personal or lightly managed devices and don't need cross-platform or mobile coverage.
Option 3 — Use a Browser-Based Translation App (Any Platform)
A browser-based translation app runs in a separate browser tab alongside your meeting. You share the meeting tab's audio to the app — using the browser's built-in tab-sharing capability, not a screen recorder or bot. The app streams real-time transcription and translation into its own tab. No one else in the meeting sees it, and nothing joins the call.
Maria is a project manager in São Paulo running weekly standups with engineering teams in Warsaw and Singapore. She can't standardize on one platform — one team uses Google Meet, the other prefers Zoom. She opens MirrorCaption in a second Chrome tab, selects Portuguese as her target language, and starts the meeting. Every spoken word from both teams streams into her MirrorCaption tab in Portuguese within half a second. No bot joined the call. The Warsaw team sees no recording notification. The Singapore team doesn't know she's reading translated captions in real time.
How to set up MirrorCaption for any meeting
- Open mirrorcaption.com in a new tab in desktop Chrome or Microsoft Edge.
- Sign in or create a free account. No credit card required — your account starts with 1 free hour of hosted transcription, one-time, no monthly reset.
- Select your source language (the language speakers will use) and your target language (the language you want to read). Choose from 50+ selectable languages.
- Click Start Session and choose Meet mode.
- When prompted, click Share tab audio and select the browser tab where your Zoom, Teams, Meet, or Webex call is running. This shares the tab's audio — not your screen — to MirrorCaption.
- Switch back to your meeting tab. MirrorCaption runs in the background. Switch to it any time to read live captions alongside the running AI-generated translation.
For face-to-face conversations — doctor visits, sales meetings, travel — use Talk mode instead: open MirrorCaption in Chrome on your phone, select your languages, and tap Start. The phone microphone captures the conversation and streams real-time translation to the screen. Hand the phone to the other person, or prop it on the table between you.
1 free hour included. No credit card. No extension to install. Works in Chrome and Edge.
Which Method Should You Use?
The right tool depends on which platforms you use and what constraints you're working under. Here's a direct comparison:
| Method | Works on | Languages | Real-time? | Bot / extension? | Free option? |
|---|---|---|---|---|---|
| Zoom AI Companion | Zoom only | Zoom-supported list | Yes | No (host setting) | Enterprise plan or add-on |
| Google Meet captions | Google Meet only | Google-supported list | Yes | No | Basic captions free |
| Teams Premium | Teams only | ~40 languages | Yes | No | ~$7/user/month add-on |
| Browser extension | Chrome-based meetings | Varies by tool | Partial | Extension required | Often a free tier |
| MirrorCaption | Any browser-based + in-person | 50+ selectable | Yes (<500ms) | No | 1 hr free (one-time) |
MirrorCaption Premium is €99 one-time — permanent product access, all future updates included, and 200 hours of hosted transcription credit included up-front. When those hours run out, Voice Packs top up at €2.99 per 5 hours; Premium users get the lowest per-hour rate. Annual plan is €54.99/year with 100 hours of hosted credit included. There is no unlimited-use tier — see the full pricing page for details.
For a deeper breakdown against Zoom specifically, see our MirrorCaption vs Zoom AI Companion comparison. For a roundup of all the main tools, see the best meeting translator 2026 guide.
50+ languages. No extension. No bot. One browser tab.
Does Real-Time Translation Work for Non-English Meetings?
Most guides assume an English speaker needing translation from one other language. Real multilingual meetings are messier: a call where some participants speak Mandarin, others speak Spanish, and the nominal meeting language is English — but nobody's first language is English.
Kenji is a Tokyo-based technical lead on a cross-functional call with counterparts in Berlin and Mexico City. The call runs nominally in English, but three people are code-switching constantly. His Berlin colleague uses German idioms when frustrated. His Mexico City colleague switches to Spanish when explaining a technical nuance quickly. Kenji misses both. With MirrorCaption open in a second tab set to Japanese output, he reads every speaker's contribution in Japanese — regardless of what language they chose to say it in.
The tools that handle multilingual calls best are the ones that let each participant choose their own output language independently, rather than requiring a host to set one translation pair for everyone in the room. MirrorCaption runs per-session and per-user: your Shanghai developer reads in Simplified Chinese, your Berlin PM reads in German, and your São Paulo CS rep reads in Portuguese — all from the same call, simultaneously, with each running their own MirrorCaption tab.
Translation quality for tonal and logographic languages — Mandarin, Japanese, Korean, Hindi — depends on the approach. Context-aware translation feeds the surrounding conversation into each translation call, which changes the output meaningfully. "少し考えさせてください" rendered word-for-word becomes "Please let me think a little." Rendered with conversational context, it reads as the polite near-refusal it is. For more on accuracy across language pairs, see our real-time translation accuracy breakdown.
Privacy: Is Your Meeting Audio Stored?
This question matters most in corporate environments, where IT and legal teams review meeting tool policies carefully. The answer depends entirely on which approach you use.
When a 200-person fintech blocked meeting bots from their corporate network, every bot-dependent transcription tool — including services that auto-join with an email address — stopped working overnight. MirrorCaption continued to work because it never joined the meeting. It captures browser tab audio using the browser's native tab-sharing API. No external process appears in the meeting participant list. No recording notification fires for other participants.
Here's how each method handles your audio:
- Platform built-in (Zoom, Meet, Teams): Audio stays within the platform's infrastructure. Zoom, Google, and Microsoft's respective data retention and privacy policies apply — your IT team likely already has these on file.
- Browser extensions: The extension vendor receives your audio stream for processing. Privacy policies vary significantly; review them before using any extension on confidential calls.
- MirrorCaption: Audio streams from your browser directly to the MirrorCaption real-time transcription layer and is discarded after processing. No audio is stored on MirrorCaption's servers. Transcripts are saved locally in your browser's IndexedDB storage — you own them and they don't leave your device unless you export them. For a full breakdown of what MirrorCaption stores, see our AI meeting privacy guide.
Frequently Asked Questions
Does Zoom have real-time translation?
Yes. Zoom offers real-time translated captions for eligible Zoom Workplace Enterprise plans and as a Translated Captions add-on for other paid Workplace plans. The account owner or admin must enable captions in Zoom account settings, and translation availability depends on the host account. It only works inside Zoom — if your meeting moves to Teams or Google Meet, you need a separate tool or a cross-platform app like MirrorCaption.
Is there a free real-time meeting translator?
Yes. MirrorCaption starts with 1 free hour of hosted transcription — one-time, no credit card required, and it doesn't reset monthly. Google Meet provides free live captions to Google account users, with translated captions available on eligible Workspace tiers. Microsoft Teams includes basic live captions; translated captions require Teams Premium or Microsoft 365 Copilot, and Teams Premium is listed at approximately $10 per user per month.
Can I translate Google Meet without an extension?
Yes — two ways. First, Google Meet's own built-in captions (and translated captions on eligible Workspace plans) require no extension. Second, MirrorCaption's Meet mode captures Google Meet tab audio in desktop Chrome or Edge without any extension, plugin, or bot. Open MirrorCaption in a second tab, share the Meet tab's audio when prompted, and read live translated captions alongside your call.
How accurate is AI meeting translation?
Accuracy varies by language pair, speaker clarity, microphone quality, accent, and domain vocabulary. No single benchmark covers every current meeting platform and language pair, so treat live translation as a comprehension aid and test your own source and target languages before relying on it for legal, medical, or commercial decisions. MirrorCaption feeds the previous 3–5 conversation segments into each translation call for context continuity.
What languages does real-time meeting translation support?
It depends on the tool. MirrorCaption supports 50+ selectable languages, including Mandarin (Simplified and Traditional), Japanese, Korean, Arabic, Hebrew, Hindi, Russian, Portuguese, Spanish, French, German, Italian, Dutch, Polish, and others. Zoom, Google Meet, and Teams Premium each publish their own supported-language lists and plan requirements — check the exact source and target language before a high-stakes meeting.
Can I translate a face-to-face conversation, not just a video call?
Yes. MirrorCaption's Talk mode runs on mobile Chrome and uses your phone's microphone. Open MirrorCaption on your phone, select your source and target languages, and tap Start. The microphone captures the conversation and streams a real-time translation to your screen. You can hand the phone to the other person so both sides read each other's words — useful for doctor visits, rental agreements, or any in-person situation where you don't share a language.
Putting It Together
The right method depends on how you work. For teams that live entirely on one platform — Zoom-only organizations, Microsoft 365 shops, Google Workspace teams — the built-in translation is the path of least resistance. Enable it once, and it's there for every meeting.
When your meetings cross platforms, when 50+ language pairs matter, when a bot or extension isn't acceptable to IT, or when you need translation during an in-person conversation, a dedicated browser-based tool is the only option that covers all of those cases without asking you to change platforms or get IT approval.
MirrorCaption takes five steps to set up — roughly three minutes from opening a tab to reading live translated captions on your next call. The first hour is free, no credit card required. If you use more, the Premium tier is €99 one-time with 200 hours of hosted transcription credit and all future updates included. No subscription, no monthly reset, no per-seat fee.
Translate Your Next Meeting in Real Time
1 free hour to try. No credit card. No bot. Works on Zoom, Teams, Meet, Webex, and in person.
Open MirrorCaption Free