AI Podcast Clip Maker: Turn Episodes into Shorts

That depends on the episode you upload, and it is worth being straight about it. Anything the AI generates for your clip is yours to use commercially — the AI voiceover, the written script, the captions, and any AI-generated or licensed stock footage that gets added. None of that carries a third-party claim, and there is no attribution requirement.

Your source recording is a different question. The clipper works from a file you provide, so if you upload your own podcast, your own stream or your own interview, the clip is yours. If you upload someone else's episode, a movie or a broadcast, clipping it does not change who owns it — and no tool can make that safe. Upload what you own or have the rights to use.

Manually, you scrub the recording for a moment that stands on its own, trim it, crop it to a vertical shape, caption it, and export — roughly ten to twenty minutes per clip, most of it spent hunting rather than editing. The hunting is the actual bottleneck, which is why a two-hour episode usually produces zero clips instead of ten.

The shortcut is to let the transcript do the finding. Upload the episode, let it be transcribed, and have the complete thoughts surfaced with timestamps — then you are choosing between candidates instead of searching a waveform. In VdoZen that is the default path: upload, review the suggested moments, render the ones you want.

No — and that is deliberate. You upload the file. There is no paste-a-URL box, because pulling video off someone else's channel is exactly the thing that gets both the user and the tool into trouble.

It works the other way round: clip the episode you are publishing, then send the clips to YouTube. You can generate a title, description and tags for each clip and publish or schedule it straight to a connected channel.

All of it, in one pass. A finished clip can carry burned-in captions in the style you pick, background music, and either the original audio or an AI voiceover written for that specific moment.

The voiceover is the part worth knowing about. It is not one flat read from the start of the clip — the narration is broken into timed lines you can move, using inline markers in the script itself. Shift a marker and the line lands somewhere else; rewrite the sentence and the timing stays put. So the narration tracks the footage instead of drifting away from it, and you set that up before paying to render.

Not for long, and deliberately not. Large source uploads are held briefly so you can re-plan and pull different clips without uploading the episode again, and then they are cleaned up automatically — within a day at the outside. Pressing "Start over" deletes the source immediately, because that is the one moment you have actually said you are done with it.

The practical effect: come back to a session much later and the source may be gone. Your browser keeps its own copy for about a day, so it is usually restored with one click; if it isn't, you'll be asked to upload the episode again, and a large file takes a minute to go back up. Finished clips are not affected — those stay available at their own links. Small files are not kept server-side at all, which is why they are the ones most likely to ask for a re-upload.

MP4, MOV and WebM up to 500MB. Podcasts, interviews, streams, webinars, talking-head recordings — anything where the value is in what is being said, which is the format this is built for. There is no fixed limit on how many clips one episode produces; a long episode simply has more in it.