For churches, meetings & classrooms

One voice. Every language. Live.

LiveTranslate listens to whoever's speaking and casts live, translated captions to every phone in the room — in each person's own language, the moment the words are spoken. No cloud service in the middle. No subscription.

Free tier, no time limit · runs on Windows · upgrade anytime to unlock full functionality

Live preview — what runs the sound booth
localhost:5000/admin LIVE
00:00:00 AM 🇬🇧 EN 98%
Live preview — what the audience sees, on their own phones
One-time purchase

Pay once, use it every week. No per-seat fees, no monthly bill to remember.

Fully offline

After setup, nothing spoken in the room is ever uploaded, streamed, or logged externally.

46+ languages

Any language Whisper understands, translated automatically to whatever your audience needs.

How it works

Three steps. No training required.

Nobody at the sound booth needs to be technical — the operator just presses Start.

01

Speak naturally

The speaker just talks, into whatever microphone is already plugged in. No special phrasing, no button to press per sentence.

02

It's heard, then translated

Speech becomes text, then gets translated into every language your audience needs — worked out entirely on this computer, offline.

03

Captions appear on their phone

Each person scans a QR code once, picks their language, and reads along on their own screen. Nothing to install.

Built for live events

Everything a live, multilingual room actually needs

Not a generic transcription tool — the details are tuned for services, sermons, meetings, and classrooms alike.

Any language, live

Speech recognition covers 46+ languages, and translation works between any two you've installed — automatically, via an English pivot.

Genuinely offline

No cloud service is involved while your event is live. Internet is only needed for setup and the one-time license activation.

Live event controls

Mic gain, pause sensitivity, and custom vocabulary — names, places, liturgical or technical terms — are all adjustable mid-event, no restart needed. An optional profanity filter keeps captions clean when that matters to you.

Take the transcript home

Every session is saved automatically and exportable as plain text, subtitles, a spreadsheet, or a Word document.

Share with a glance

A QR code on the screen is the whole onboarding flow. Scan it, pick a language, done — no app, no account.

Put it on the big screen

An optional display window drags onto a projector or second monitor, showing large, room-scale captions in up to four languages at once — separate from the admin panel and everyone's phones.

Proof, not a promise

What's actually on screen during a live session

The preview up top is built to load fast on this page. Here's what the real admin panel and the real phone view look like — drop in your own captures below.

Screenshot needed
assets/screenshots/admin-panel.png
Admin panel, mid-session — the operator's screen
Screenshot needed
assets/screenshots/phone-view.png
One phone in the audience, captions live
Screenshot needed
assets/screenshots/display-window.png
Display window, projected — large captions for the room
Recording needed
assets/screenshots/walkthrough.mp4
30–60s walkthrough, start to finish
Social proof

Be the first voice here

LiveTranslate is new enough that this space is still empty, on purpose — we're not going to put words in anyone's mouth.

Run an event with it, then tell us how it went. The first few real quotes go right here.

Pricing

One license. One computer. No subscription.

Start free, no time limit — upgrade whenever you're ready. Everything you set up carries over.

Do the math

A subscription captioning service charges every month, forever. LiveTranslate is €50, once. See where the lines cross.

€50 LiveTranslate, total — ever
€300 A subscription, over 12 months

That's €250 more than LiveTranslate costs, total — and the subscription keeps billing after month 12.

LiveTranslate €50one-time

Download and use LiveTranslate free, with no time limit, though some functionality is limited. A single license unlocks full functionality on one computer, permanently.

  • Full functionality — no limits
  • All 46+ languages included
  • Runs fully offline once set up
  • Optional GPU acceleration, same license
  • Transcript export — txt, srt, csv, Word

Local taxes may apply and are calculated at checkout. Payment processed securely by Lemon Squeezy.

Questions

Before you press Start

Only twice: once during setup to download language packs and the speech model, and once to activate your license key. After that, it runs completely offline — including during a live event.

Any Windows PC works — a modest one runs the smaller speech model comfortably. A PC with an NVIDIA GPU can run a larger, more accurate model with almost no lag.

No. Speech recognition and translation both run on the same PC LiveTranslate is installed on. Nothing is uploaded, streamed, or logged externally.

The free tier has no time limit and no credit card required, but some functionality is limited. A one-time license removes those limits and unlocks full functionality on that computer, permanently. Nothing about your setup is lost when you upgrade — settings and saved transcripts carry over the moment you activate.

Choose any of the 46+ languages during setup. You can add more later at any time by re-running the installer.

No. Setup asks whether you're running a church service or a general meeting, conference, or classroom, and starts you off with sensible defaults either way — mic sensitivity, vocabulary hints, and the profanity filter. Everything stays fully adjustable afterward regardless of which you pick.

Yes. A separate display window can be dragged onto a projector or second monitor, showing large captions in up to four languages at once — alongside the phone captions, not instead of them.

[Placeholder — confirm and fill in: is there a self-service way to deactivate a license on the old machine and activate it on the new one, or does this need to go through support?]

LiveTranslate runs on Windows 10/11. The smaller speech model runs comfortably on modest, several-years-old hardware; the larger, more accurate model benefits from more RAM and, optionally, an NVIDIA GPU. [Placeholder — confirm and fill in: minimum/recommended RAM, free disk space for language packs + model, and GPU VRAM for the large model.]