An AI video editor that cuts your takes and animates what you say

How to edit a talking-head video with AI

You bring the footage. The AI editor cuts it, the AI director animates it, and you keep the last word in the studio.

1. Upload your clips: MP4, MOV or WebM, one or several, up to 5 GB each, up to 20 clips and 30 minutes in all. Put them in the order they should play.
2. Choose the voice. Keep my voice makes what you say the narration. AI narrates writes a new one from your clips, read by a voice you pick.
3. Leave on the switches you want: Keep the best takes, Cut silent moments, Camera moves, Replays & slow motion. If you spoke from a script, paste it in Script & notes.
4. Pick the style and the shape, and click Generate. The price is shown first, and no credits are used before that click.
5. Watch the edit in the studio. Trim, split or pace any scene, click one to change it, or ask the video assistant in plain words. Export a 1080p MP4.

An AI that cuts your bad takes: what “Keep the best takes” does

Say you film your course intro on a Sunday afternoon. The opening line comes out right on the fifth try. You lose the thread halfway, start again, ask someone off camera whether the light looks fine, and say the last part twice because you weren’t sure. That is normal filming. It is also an evening of work on a timeline.

A professional editor gets through it in passes: watch everything, group every delivery of each line, keep one take per line in the script’s order, then fine-tune each cut so no word is clipped. Keep the best takes runs the same passes. The AI editor hears every take and sees what the camera saw, then keeps the one that is complete, said right, well delivered and looking at the lens.

It cuts false starts, stumbles, repeats and off-camera talk, follows your script’s order when you give it one, and evens out the sound between clips so one take isn’t louder than the next. It leaves a few natural pauses and fillers in, because a voice with every breath removed sounds robotic. It chooses among the takes you filmed: it doesn’t build sentences you never said out of pieces of different ones.

In one real course intro, 29 minutes of raw takes became a 3-minute intro. Read how that edit went.

How do you add motion graphics to a talking-head video?

In the Motion style, the AI director builds the graphics around you, beat by beat, from what you say. Nothing comes from a template: each video gets its own layouts and its own timing.

• Your video becomes a window that moves: full screen while you speak to the viewer, a card beside big words when you make a point, a small window in a corner while a full-screen animation plays.
• A name card the first time you appear: your name and your role, on the side away from your face.
• Big words beside you on the line that matters, never over your face.
• Frosted glass panels for lists: say “here is what you will learn” and the list appears, each item lit on the word that starts it.
• Pictures as cut-aways while your voice carries on: from your Library, from your uploads, or made by AI for that line.
• Full-screen animation when an idea needs the whole frame, then back to you.

The rules behind it are old ones. Richard E. Mayer’s principles of multimedia learning ask for signaling (point at what matters), coherence (leave out what doesn’t) and temporal contiguity (show a thing when its word is said). So a picture lands on the word that names it, one main element holds the screen at a time, and on a personal line the director leaves your face alone.

Camera moves, replays and silences: the rhythm of the edit

An edit lives on its rhythm. Three switches handle it, all on by default, and a fourth adds sound.

• Cut silent moments: the pauses where nobody speaks are taken out, so the video keeps moving.
• Camera moves: the AI zooms and pans on your footage where it helps, a slow push, a quick zoom on a detail. Every cut between two takes is hidden: the frame jumps a little closer or wider right on the cut, as if a second camera had filmed you, or a picture or a card covers it.
• Replays & slow motion: a quick moment, a gesture, a click, shown again slowed down to explain it. With Keep my voice, your voice pauses while it plays.
• Sound effects, a switch at the last step: the AI designs whooshes, hits, clicks and the sound of the place, timed to what moves. No music. You see, move and change every sound on the timeline.

Keep my voice or AI narrates?

Keep my voice is for when you talk to the camera. What you say in your clips is the narration, and the motion graphics follow your words.

AI narrates is for footage where your face isn’t the point: a screen, a product, hands at work, a drill. Clipteach writes a new narration from what you say, or from what your clips show when you don’t speak, read by a voice you pick. Even over a silent clip, the narration teaches the viewer what to do and what to watch for; it doesn’t just describe the screen.

Script & notes, beside your clips, is optional. Paste the script you spoke from, who says each line, names spelled right, and notes for the edit, like “show the roadmap screenshots here”. The AI follows it; your recording stays the voice.

What kinds of videos can you edit with it?

• Course intros and welcome videos: the first thing a new student sees, cut tight, with your name on screen and what they will learn. More for course creators.
• Explainers and lessons filmed to camera, with the key points written beside you.
• Talking-to-camera videos, the YouTube “yapping” kind: one person, one idea, paced like an edited video.
• Tutorials: you explain, the camera closes in on the detail, a replay shows the quick step again.
• Announcements: a launch, an update, a new offer, in 16:9 for your site and in 9:16 for social. More for marketing teams and founders.

Two people talking? The podcast video editor tells the voices apart and keeps the camera on whoever speaks. Filming a drill or a gym set? The sports video editor follows the body and, with 3D Body, paints the muscles at work. No footage at all? Start from text, a document or your voice.

16:9 or 9:16, and what it costs

Your video comes out as a 1080p MP4: 16:9 for YouTube, courses and websites, or vertical 9:16 for Reels, Shorts and TikTok. It starts in your first clip’s shape, and you can switch it at the Format step.

Your clips are prepared when you click Generate, and no credits are used before that. The price is shown first, based on the finished video’s length.

You start with 300 free credits on sign-up, no card, enough for a short video. After that, Starter is $19 a month, Creator $49 and Studio $149, and credits roll over and never expire.