How podcast to video works
Bring the recording you already have. Four steps, and the studio is yours.
- Upload the episode: MP4, MOV or WebM, one clip or several in the order they play, up to 5 GB each and 30 minutes in all.
- Choose Keep my voice, and leave Keep the best takes and Cut silent moments on. In Script & notes, say who is who and spell the names, for example [ANNA] and [TOM] before their lines.
- Pick the Motion style and the shape, 16:9 or 9:16, and click Generate. The price is shown first, and no credits are used before that click.
- Check the edit in the studio: trim, split, change any scene or ask the video assistant. Export a 1080p MP4.
Who is speaking? Clipteach works it out
You don’t tag speakers by hand. Clipteach tells the voices apart, then looks at a few frames to match each voice to a face. From then on the AI director knows who says each line and where that person sits in the frame.
Names come from your Script & notes or from what people say on the recording (“I’m Anna, and this is Tom”). The AI never invents one, so if nobody says their name, write it in the notes.
Each person gets a name card for a few seconds the first time they speak, on their side of the frame, near them.
Interview video editing: the camera goes to whoever is talking
A conversation filmed in one wide shot is hard to watch for long: two small faces, and nothing moves. A human editor fixes it by cutting between angles. Clipteach builds those angles from the one shot you have.
- When someone says a line that matters, the camera pushes in on their face, then goes back to the wide shot when you trade lines.
- Or the frame splits into two windows side by side, one face in each, the speaker’s window lit by its outline.
- In a vertical video, the two windows stack one above the other.
- Each framing holds for a couple of seconds. It doesn’t cut on every “yeah”, so the edit never feels like a tennis match.
- A line you say together shows you together.
Film everyone in one wide shot, with each face clear and lit. Clips play one after another, so two cameras that recorded at the same moment aren’t synced; the AI frames each face from the wide shot instead.
AI podcast editing that keeps the conversation
AI podcast editing often means stripping out every “um” until the voices sound pasted together. Keep the best takes edits the way a professional does: listen to everything, keep one good version of each line, fine-tune every cut. On a conversation it also keeps what makes it a conversation: the laugh after a joke, the “right” that answers a point.
False starts, stumbles, repeats and off-camera talk come out. With Cut silent moments on, so do the pauses where nobody speaks.
The sound is evened out between clips, so one recording isn’t louder than the next, and the Voice volume control in the studio raises or lowers the whole voice track.
What to film, what to leave to the edit, and how one episode becomes several videos: podcast to video, step by step.
Big words and pictures on the key points
Most of a podcast should be the people. But when a guest gives the number, the name or the claim the episode is about, it deserves to be seen.
- Big words beside the speaker, on the line that matters, away from their face.
- A picture as a cut-away, or in a card beside them, while the voice carries on: from your Library, your uploads, or made by AI.
- A frosted glass panel when someone lists points, each one lit as it is said.
- Full-screen animation for an idea that needs the whole frame, in the Motion style, then back to the two of you.
The director changes something on screen every few seconds, and only when it means something: a new point, a number, an example. On a personal story, it leaves the faces alone.
A vertical clip for Reels, Shorts and TikTok
Choose vertical 9:16 at the Format step and the same conversation is framed for a phone: when both speakers are on screen, their two windows stack one above the other, one face in each, and the words stay clear of the buttons Reels, Shorts and TikTok put on the screen.
For a short clip, upload only the minutes you want to post. Each video is priced on its own length, shown before you start. You begin with 300 free credits on sign-up, no card.
What the podcast editor doesn’t do
Better to know before you upload:
- It doesn’t write out every word along the bottom of the screen. Words appear on screen as big words on the key points, not as a running transcript.
- It doesn’t add music. Sound effects are an optional switch, and the AI designs them as whooshes, hits, clicks and the sound of the place, never a soundtrack.
- It doesn’t sync several cameras that recorded at the same moment: clips play one after another. Film one wide shot, or upload the multicam edit you already made as one file.
- It doesn’t match colour between clips, and it adds pictures, not generated video.
- It takes up to 30 minutes of footage per video. For a longer episode, choose the part you want, or make several videos.
- No avatars, no stock footage: the people on screen are the people you filmed.
One person talking to camera? The AI video editor covers course intros, explainers and tutorials. Turning episodes into videos for a course or a product? See Clipteach for course creators and for marketing.