Bilingual sermon translation
Switch languages mid-sentence. Everyone still hears their own.
Voclyn detects the language of every phrase, even mid-sentence, so nobody sets a language at the sound desk. A pastor can tell the story in Korean and make the point in English; every phone keeps up.
It hears which language is being spoken, and each person reads the service in their own.
See it
Every speaker in their own language.
The pastor, an elder and the reader each speak their own language, and every pew follows in its own. Nobody touches a setting. Press play.
At the pulpit
Good morning, church. Welcome to our joint service — three congregations, one family.
In the pew
Buenos días, iglesia. Bienvenidos a nuestro culto unido: tres congregaciones, una sola familia.
One service instead of two
A church that worships in two languages usually holds the same service twice, or preaches through an interpreter. With an interpreter every sentence is said twice, so a 45-minute sermon becomes 90 minutes, and only one other language fits.
With Voclyn the pastor preaches once, at the usual pace, in whatever language the room speaks, and mixes two in one sentence if that is how the congregation talks. Each person follows on their own phone, reading or listening in their own language. How live sermon translation works.
How Voclyn hears a change of language
The speech engine marks every word with the language it was spoken in. When the language changes, Voclyn ends the phrase there and sends it on, so each part is translated from the language it was actually spoken in. Nobody presses anything at the sound desk.
It hears the languages churches preach in, among them English, Spanish, Cantonese, Mandarin, Korean, Vietnamese, Tagalog, Portuguese, French, Persian and Arabic. All 100+ languages.
In Settings, under Spoken from the front, a church can leave Auto-detect on or name the languages usually spoken. Naming every language that might be spoken, even briefly, makes recognition more accurate. Name Cantonese if it’s spoken, since Cantonese gets its own speech engine.
What each person reads
Each phone shows every sentence in its own language. When the speaker is already using a listener’s language, that listener reads the words as spoken, marked Original. Everything said in other languages arrives translated, with a small tag naming the language it came from.
| Said at the front | An English phone reads | A Spanish phone reads |
|---|---|---|
| Dios nos ama | God loves us. (from Spanish) | Dios nos ama. (Original) |
| every single day | Every single day. (Original) | Cada día. (from English) |
| sin excepción. | Without exception. (from Spanish) | Sin excepción. (Original) |
One sentence from our test on 29 October 2026, as it reached the phones: Spanish, then English, then Spanish again.
With the AI voice on, the voice reads the translated parts and goes quiet while the speaker uses the listener’s own language; the page says “Original · you’re hearing the speaker”. About the AI voice.
Several voices, each told apart
A service is rarely one voice: a pastor, an interpreter, a reader, an elder praying. In rooms that don’t include Cantonese, Voclyn tells the voices apart, and each speaker’s sentences get their own colour rail on the listener’s page, labelled Speaker 1, Speaker 2. A voice keeps its colour when it changes language. How speaker detection works.
What we tested on 29 October 2026
We streamed sentences in English, Spanish, Mandarin and Korean through the live system, two of them switching language midway, and timed what phones reading English, Spanish and Mandarin received. Every part was labelled with the right language, and the two mixed sentences reached the phones as quickly as the rest.
Across the whole test, each sentence was translated right after it ended. How we test.
Honest limits
A sentence that switches midway arrives in pieces
Each part is translated as soon as it ends, so other languages read a mixed sentence as two or three short lines, and a part can read unfinished on its own.
Said at the front
God’s love is for 每一个人 everyone here.
A Spanish phone, in three pieces
El amor de Dios es para
Cada persona.
Todos aquí.
Cantonese rooms
Rooms that include Cantonese use their own speech engine, which our test didn’t cover, and they don’t show speaker colours yet.
Questions
What churches ask.
Something else on your mind? Write to us at hello@voclyn.com.
The languages churches preach in, among them English, Spanish, Cantonese, Mandarin, Korean, Vietnamese and Tagalog, each detected by itself. People in the pews can read 100+ languages.
No. Voclyn detects the language of every phrase by itself, and nothing changes at the desk when the speaker switches. In Settings you can leave Auto-detect on or name the languages usually spoken, which makes it more accurate. Name Cantonese if it's spoken, since it gets its own speech engine.
The words as spoken, marked Original. Whatever is said in other languages is translated into theirs, and the page says AI translation. With the AI voice on, the voice goes quiet while the speaker uses their language.
Yes. Voclyn captions the interpreter like any other voice and, in rooms without Cantonese, gives each voice its own colour. Today the interpreter's words are translated for everyone too, so people read each sentence twice. Interpreter mode, for routing an interpreter's words, is coming next.
Not in our test on 29 October 2026: the two sentences that switched language midway reached the phones as quickly as the rest.
Yes, Free included. Free translates 6 hours a month into one language plus your main language; paid plans have unlimited translation hours (fair use).
Preach the way your church talks.
Try it in the demo: speak two languages in one sentence and watch the big screen and two phones follow. 30 days of the Church plan free, no card.
Your first month is on us: the Church plan, normally $99 a month, with unlimited listeners, translation hours and AI voiceover. No card; then choose a plan, or stay on Free.