Live Translation for
Online Tutoring

MirrorCaption gives online tutors real-time transcription and translation in 50+ languages, delivering captions in under 500ms — fast enough for students to read while you're still speaking. It works alongside browser-based Zoom, Google Meet, Microsoft Teams, and Webex, with no bot joining the meeting and nothing to install on either side.

It's Tuesday afternoon in Seoul. Your student has paid for 50 minutes of English conversation practice. The lesson is going well until you introduce the subjunctive: "If I were you, I'd apply now." You watch the familiar expression cross the screen — polite nodding, glazed eyes, zero comprehension. You slow down, repeat, rephrase. The lesson stalls for four minutes. Then five.

Now imagine that same student reading a Korean translation of every word you say, word by word, as you say it. The lesson doesn't stall. The comprehension gap doesn't open. You both keep moving.

Why Real-Time Translation Changes How You Teach

Every comprehension gap that goes unaddressed mid-lesson costs time. A student who doesn't understand "subjunctive" at minute 12 of a 50-minute session doesn't just miss that concept — they lose the thread for the minutes that follow. By the time they raise a hand, you're on a different point entirely.

Post-session tools don't solve this. Tools like Otter.ai produce excellent summaries after a call ends. For reviewing vocabulary or revisiting an explanation, they're genuinely useful. But they arrive after the lesson is over. The comprehension failure happened at minute 12. The transcript arrives at minute 51. The gap doesn't close.

What actually breaks comprehension mid-lesson

The problem isn't effort. Spoken language moves faster than the brain can parse an unfamiliar word, infer meaning from context, and continue tracking the next sentence — especially in a second language. When that processing stalls, the student falls behind. And once they're behind, they stay behind until the tutor notices, or until they interrupt.

Neither outcome is ideal. The tutor loses flow. The student loses face. Both lose time.

Real-time translation removes the bottleneck. A Korean student reads "가정법" (subjunctive mood) alongside your English explanation — not as a replacement, but as a scaffold. They understand the English faster because the Korean is there when they need it. The lesson keeps moving.

Comprehension in the moment, not after it

Priya has tutored SAT Math online for three years. Most of her students are in China; their English is strong for reading but struggles with spoken mathematical instructions. "Differentiate the expression with respect to x" arrives as a wall of sound. By the time they've parsed it, Priya has moved on.

In this illustrative tutoring workflow, Priya stops losing 8–10 minutes per session to re-explanations because students can catch missed phrases in the side-by-side transcript without interrupting. That is what in-session comprehension does — it compounds.

How MirrorCaption Works in an Online Tutoring Session

There's no new app on your student's side, no calendar connection, and no meeting bot to invite. The workflow takes under two minutes to start:

1. Open MirrorCaption in your browser — use desktop Chrome or Microsoft Edge for meeting sessions (Meet mode). Your student can do the same on their own device, independently and simultaneously.
2. Select Meet mode and share your meeting tab — MirrorCaption captures what's said in the meeting tab through the browser's audio APIs. No plugin, no screen-capture tool required.
3. Choose your language pair — set the source language (the language being spoken) and the target language (what the student reads). Tutor and student can set their own preferences independently.
4. Teach as you normally would — captions stream word by word as you speak, appearing in the student's chosen language with under 500ms end-to-end latency. No mute buttons, no pauses, no "processing..." delay.

The student doesn't need host permissions or a paid platform account. They open MirrorCaption in their own browser tab and read the live transcript independently. For in-person tutoring in the same room — different first languages — open MirrorCaption in Talk mode on a phone in mobile Chrome and place it between you. As you speak, the student reads in their language.

Features That Make the Difference in Tutoring

MirrorCaption was built for bilingual conversations, not for passive meeting recordings. Several features become especially effective in a tutoring context.

Word-level tap-to-original

Every translated word is linked back to the exact source word it came from. Tap any word in the translation and MirrorCaption shows you the original. For language-learning students, this is significant: they can check whether "application" was translated as "申請" or "応用" and develop intuition for which word fits which context. The translation doesn't hide the source — it reveals it.

This is a different approach from tools that replace the original entirely. MirrorCaption keeps both visible. Students build vocabulary from the connection between them.

Vocabulary Builder — every lesson becomes a study deck

While the transcript streams, students can tap any word or phrase to save it to their personal Vocabulary Builder. By the end of a 50-minute session, they've built a deck from the actual language of the lesson — not constructed textbook examples, but real English sentences in context. This is language learning with real meetings, not simulated practice.

Tutors teaching English conversation can encourage students to save 5–10 new words per session. The list becomes self-generated homework with no extra preparation from either side.

Side-by-side view — original and translation together, not instead

MirrorCaption displays the original transcription and the translation side by side on desktop, not stacked or alternating. The student reads both simultaneously. For language tutors teaching Japanese, Chinese, or Spanish to English-speaking students, this matters: the student sees "仮定法" alongside "subjunctive mood" and builds a real connection between the two, rather than relying on the translation alone.

Tools designed purely for comprehension speed replace the source with the target. Side-by-side is designed for comprehension and acquisition at the same time.

Speaker detection

MirrorCaption automatically detects distinct voices and labels them. Both participants can rename the labels (Tutor / Student, or their actual names). In the transcript, it's clear whose words are whose — useful for students reviewing a session afterward, and for tutors who export a verbatim record of what was explained versus what was asked.

Works With Your Tutoring Platform

MirrorCaption isn't tied to any single meeting tool. It captures meeting-tab audio from whatever browser-based video call you're already using:

• Browser-based Zoom
• Google Meet
• Browser-based Microsoft Teams
• Browser-based Webex
• In-person sessions via Talk mode on mobile Chrome

No integration setup. No calendar connection. No bot to invite. Whichever platform your tutoring marketplace or student prefers, MirrorCaption works alongside it without touching the meeting itself — just like having a personal subtitle track in your browser tab.

This is meaningfully different from platform-native translation tools. Zoom translated captions require Zoom Workplace Business Plus, qualifying Enterprise plans, or a Translated Captions add-on per Zoom's support page; Google Meet's translated captions require a qualifying Google Workspace plan; Microsoft Teams live translation is part of Teams Premium. All three are locked to their respective platform and account setup. MirrorCaption works for any tutor, on any plan, using any browser-based meeting tool. See how it compares: MirrorCaption vs Zoom AI Companion.

No bot joins your session. In a 1:1 tutoring call, a bot presence is immediately visible to the student. It changes the atmosphere. MirrorCaption captures audio locally in your browser tab through standard browser APIs. Other participants see nothing added to the meeting — no extra name in the participant list, no notification. For multilingual remote teams and for tutors alike, that discretion matters.

Languages Supported — 50+, Including the Top ESL Pairs

MirrorCaption supports 50+ selectable languages. You choose the source and target before each session, and you can change the pair mid-session. The most common tutoring pairs:

• English ↔ Mandarin Chinese (simplified and traditional)
• English ↔ Japanese
• English ↔ Korean
• English ↔ Spanish
• English ↔ Portuguese (Brazilian and European)
• English ↔ Arabic
• English ↔ Hindi
• English ↔ French
• English ↔ German
• English ↔ Russian
• English ↔ Turkish
• English ↔ Indonesian

Language pairs run bidirectionally — if you're an English speaker learning Japanese with a native tutor, set Japanese as the source and English as the target. The tutor's words appear in Japanese; the translation column shows English.

A real example: A tutor explains the Japanese concept of "曖昧" (deliberate ambiguity in communication — a culturally loaded word that reflects a broader social norm). An English-speaking student reading the translation column sees "ambiguity." That's technically correct. But with side-by-side view and tap-to-original, they see "曖昧" in the source column and can investigate the nuance — something a pure translation pass would erase.

Who Uses MirrorCaption for Tutoring

Pricing — Start Free, Decide Later

MirrorCaption includes 1 free hour of hosted transcription — one-time, no monthly reset, no credit card. Open the app, run your next tutoring session, and see whether it fits before committing to anything.

Voice Packs are sold separately on every plan and cover hours beyond the included credit. Premium customers get the most favorable rate: 5 hours for €2.99 (€0.60/hr), 15 hours for €7.99 (€0.53/hr). No subscription required to top up.

Premium at €99 one-time is permanent product access — no recurring billing, no plan expiry, all future product updates included. It doesn't include unlimited transcription time: the 200h credit covers most tutors for many months, and additional hours come from Voice Packs at the lowest available rate.

For a tutor teaching 10 hours a week, the €99 Premium pays for itself in well under two months compared to tools priced at $16–20/month with real-time translation locked behind higher tiers. See how pricing compares across tools in our best meeting translator 2026 roundup.

Frequently Asked Questions

Does MirrorCaption work with Zoom tutoring sessions?

Yes. MirrorCaption works alongside browser-based Zoom in desktop Chrome or Microsoft Edge. Open MirrorCaption in a separate tab, select Meet mode, share your meeting tab audio when prompted, and live captions appear in your chosen language. You don't need to be the meeting host or have a paid Zoom plan — MirrorCaption captures audio on your side of the call independently.

Can my student use MirrorCaption without installing anything?

Yes. MirrorCaption is a browser-based web app. There's nothing to download, no browser extension to install, and no meeting bot for the student to approve. They open the URL in Chrome or Edge on their own device during the session and start reading captions independently. Setup takes about 30 seconds. They don't need a MirrorCaption account to read a shared transcript view, though creating a free account unlocks the Vocabulary Builder.

Does a bot join the tutoring call?

No. MirrorCaption captures meeting-tab audio directly in the browser using standard browser capture APIs. Nothing joins your meeting. The other participant — your student — sees no notification, no third-party name in the participant list, and no change to the meeting interface. It's a personal comprehension tool that runs in your browser tab, not inside the call.

What languages are supported for online tutoring?

MirrorCaption supports 50+ selectable languages, including Mandarin Chinese, Japanese, Korean, Spanish, Portuguese, Arabic, Hindi, French, German, Russian, Turkish, and Indonesian, among many others. You select the source language and target language before each session. Language pairs are bidirectional, and you can change the pair mid-session without restarting.

Is the free tier really free — no credit card?

Yes. Every account includes 1 free hour of hosted transcription — one-time, no monthly reset, and no credit card required to start. The free hour gives you full access to Meet mode, Talk mode, 50+ languages, speaker detection, word-level tap-to-original, and the Vocabulary Builder. Upgrade to Annual (€54.99/year, 100h included) or Premium (€99 one-time, 200h included, all future updates, lowest Voice Pack rate) when you need more transcription time.

Does MirrorCaption record or store my students' audio?

No. Audio is processed in real time in the browser and immediately discarded — nothing is stored on MirrorCaption's servers. Transcripts are saved locally in your browser's IndexedDB storage only if you choose to keep the session. Your students' words never leave your session environment. This makes MirrorCaption suitable for tutoring contexts where recording consent is a concern: it doesn't record, and it doesn't announce itself to other participants.