A sermon is not raw material. It is a message somebody prayed over and preached to people they know by name. The AI here finds, cuts, frames and transcribes what was said. It does not create anything the preacher did not. Below are the lines we drew on purpose, and separately, the limits we simply have not built past yet, so you never have to guess which is which.
We will never train a model on your pastor's voice or generate audio in it. Every second of sound in a clip was recorded in your room on Sunday. Other tools sell this as a feature; we think a congregation should be able to trust that the voice is the person.
The AI does not write a sentence, a summary line or a caption that the preacher did not say. Captions are the transcript, word for word. The Sermon Packet's notes, quotes and devotional are built from the preacher's own words, and a check in the pipeline drops any line the transcript cannot account for.
No stock B-roll, no music bed, no motion graphics over the message. A clip is the pastor, the room and the words. If the camera was on the worship team for the whole moment, the clip is dropped rather than papered over.
We do not post to your accounts, and we do not schedule anything. Clips land in your dashboard and you choose which ones go out. A clip that misses the point of the message is seen by you and nobody else.
Moments are ranked on whether they carry the sermon's main point, and a clip that does is always in the set. Openings, announcements and the closing prayer are skipped. We do not optimise for the loudest few seconds.
No made-up churches, pastors, follower counts or star ratings anywhere on this site. The numbers on the homepage are read from the database each hour. There are no testimonials yet because no church has been asked for one yet.
These are not principles. They are things that are not built. If one of them is the reason you cannot use this, tell us on the integrations page in your dashboard; that list decides what gets built next.
Captions are in the language the sermon was preached in, and we do not translate them into a second one. Spanish, Korean, Swahili and most other languages are captioned in their own words.
Every clip is vertical 9:16 for Reels, Shorts and TikTok. There is no square or landscape version of a clip today.
We do not add your church's logo, a watermark or an intro. Paid clips come out clean so you can add your own in Canva or CapCut in a couple of minutes.
It will, sometimes. Every clip in your dashboard has its own controls: move the start or end, change who the camera follows, turn captions on or off, choose how much audio clean-up to apply, and re-render just that clip, up to 3 times. If the moment you wanted is not in the set at all, pick it straight from the transcript, up to 3 per sermon while the source is stored. And every clip has a thumbs-down and a note box that a person reads every day.
The full list of what happens to a sermon, step by step, is on the how it works page.
Two sermons free, no card. Nothing is posted anywhere. You see the clips first, and you decide.
Clip a sermon free