The questions we keep getting.

About 20 minutes the first time. You create an account, name your service, paste a list of languages your congregation speaks, and connect either your phone or an audio cable from your board. We provide a step-by-step on the first call.

No. They scan a QR code or visit a short URL. The listener experience is a regular web page — no app, no login, no permissions beyond opening the browser tab.

Whatever you already preach with works. We accept input from a phone mic, a USB mic, a wireless lavalier into a phone, or a direct line out from your sound board. Cleaner audio in = cleaner translation out, but we tolerate a wide range.

Yes. Book a free live test. We’ll set you up with a temporary URL, run a real Sunday with you, and review the transcripts together afterward. Zero commitment.

On clean sermon audio we average 95–98% word accuracy in English transcription, and translation quality compares favorably to a competent human translator on a first pass. We’re honest about the failure modes: jokes, idioms, and unfamiliar proper nouns are where humans still beat us.

Currently 30+ languages, with the most coverage on Spanish, Portuguese, Mandarin, Korean, Vietnamese, Tagalog, Arabic, French, and Russian. New languages added on request — most within 48 hours.

Yes. Every account has a glossary editor. Add proper nouns, theological terms, and language-specific overrides. Glossaries apply at both transcription and translation, so "Yahweh" stays "Yahweh" across every language.

Speech-to-text handles most major accents well — Southern US, Indian English, African English, Spanglish, etc. For unusual dialects we recommend running a 10-minute test recording first; we can tune from there.

No. Your audio and transcripts are never used to train any model — ours or our vendors’. We have data-processing agreements with each transcription and translation vendor that explicitly exclude training use.

You do. Recordings live in object storage tied to your account. You can export, download, or delete them at any time. We retain only what you ask us to retain.

Compliant. We’re happy to sign DPAs. Listener identity is not required and not stored by default.

Not yet. Our roadmap includes a self-hosted appliance for organizations that require it (large denominations, government conferences). Email us if this is a hard requirement and we’ll add you to the early-access list.

Our broadcaster device caches audio locally and resumes streaming when the connection returns. Listeners see a "reconnecting" badge but their session doesn’t break. If your service has unreliable internet, we recommend our edge-cache appliance.

A moderator can pause, edit, or take the stream offline at any moment from the broadcaster console. You can also flag a transcript line after the fact and we’ll review it together to improve your glossary.

Latency, accuracy, and control. Generic tools are 5–10 seconds behind, translate word-by-word instead of sentence-by-sentence, and have no concept of your congregation’s vocabulary. They also stop where free tier stops. MOD Translate is purpose-built for live, in-the-room translation.

Fair concern. We’re profitable on church revenue alone and not chasing venture growth at the cost of our customers. If anything happens to us, your data exports cleanly to standard formats and your recordings are yours forever.