A podcast clip has a harder job than any other kind. It lands in front of somebody who has not heard the episode, does not know either voice, and is scrolling with the sound off. SnapDub cuts for that: framed on whoever is talking, captioned, and ended on a finished thought.
500 free credits every month, no card.
A vertical frame is a third as wide as the recording you made. Two people sitting across a table do not both fit in it. Something has to decide who the viewer is looking at, and it has to keep deciding, because they take turns.
SnapDub follows the person who is speaking. When the answer comes back and the other one takes over, the frame goes with it. Nobody ends up as half a shoulder at the edge of the picture, and nobody watches a listener nod for nine seconds while somebody off-screen makes the point.
Interviews do that. Where two people are clearly a pair sitting apart, the clip splits into a panel each instead of swinging between them - both faces, full width, no ping-pong. Where they are too close together to separate, it pulls back and keeps them both in frame rather than choosing wrong twice a second.
Cut a podcast on a stopwatch and you get a clip that opens mid-sentence and stops mid-word. The viewer does not know they have joined something halfway through; they just feel that it does not make sense, and they keep scrolling.
So the boundaries are put where the conversation puts them. A clip starts where a thought starts and ends where it lands, with a breath either side rather than a hard slam into speech. If the edit has a cut right there, the clip starts after it, not four frames before it on somebody else's shot.
Some answers are twenty seconds. Some are two minutes and worth every second. Setting one length for the whole episode makes half the clips feel padded and cuts the other half off at the interesting bit, so the clips come out at the length the point actually takes.
A podcast is the one format that is entirely speech, and short-form feeds are the one place people watch without sound. A clip with no words on screen is a clip of two strangers moving their mouths.
Every clip comes out captioned, word by word, in the style you picked once. Names, jargon and the odd proper noun are yours to fix in the editor, and the fix stays on the clip rather than needing a re-render.
If the episode already ships with subtitles - a file you exported, or a track inside the recording - those are used as they are. They are the real dialogue rather than a guess at it, and on an episode that is not in English they are usually better than anything a transcriber will do.
An episode goes out once. The clips from it can carry the account until the next one, which is the whole reason anybody clips a podcast in the first place.
So the batch comes back with a title, a description and hashtags on each clip, and you can post them straight to TikTok or YouTube from the same page, or line them up on a schedule and leave them. The next episode runs on the same template and comes out looking like the last one.
No. Give SnapDub the episode's video - a link or a file - and it takes the sound from it. Clips are video, so a pure audio recording has nothing to show on screen.
It follows whoever is speaking rather than naming them. For a clip that is what matters: the person talking is the person on screen, and the frame changes hands when they do.
No. Long is the normal case here - a two-hour episode is what makes a batch of clips worth having in the first place. What you spend depends on the length of what you feed in, not on how many clips you take out of it.
That happens on most remote episodes and it does not stop anything. The framing reads each speaker separately, so a softer webcam feed is still cropped and captioned like the rest.
Yes, in the clip editor, and the correction stays on that clip. Guest names and jargon are the two things worth a glance before posting.
Paste the link. The first batch tells you more than any description of it will.