Turn long videos into short clips

The problem

You already made the content. An hour-long podcast, a webinar, a recorded talk, a long client Q&A. Somewhere in it are ten moments that would work as short vertical video, which is the format that actually travels. Producing those ten clips by hand means finding the moments, cutting, reframing to vertical, adding captions and exporting, which is an afternoon per episode.

So it does not happen. The long recording gets published once to an audience that already knows you, and the reach that short clips would have brought never arrives. The bottleneck is not ideas or material, it is editing labour on content that already exists.

What AI can actually do here

This is one of the clearer wins in the AI tool landscape, because the task is well defined and tedious.

Automatic clipping analyses a long video and proposes the segments most likely to work standalone, using speech at minimum and, on paid tiers, visual objects, sound and emotional cues. It reframes to vertical and follows the speaker, generates animated captions from the transcript, and exports ready for TikTok, Instagram and YouTube Shorts.

Transcript-based editors approach the same material differently: you read the transcript, delete what you do not want, and the video follows. That is slower than automatic clipping and gives you control over exactly where a clip starts and ends.

The honest limit is judgment. Automatic selection finds moments that look like moments, high energy, a laugh, a rise in pitch, which is not always the same as the moment that carries the idea. Expect to reject half the clips and to trim the start of the ones you keep, because automatic clips very often begin one beat too early and the first second is what decides whether anyone watches.

How the tools differ

OpusClip is built for exactly this job. It finds clips, reframes them and captions them, with a free tier that exports at up to 1080p with watermarks on animated caption templates and 3-day storage. Starter at $15 per month removes watermarks, extends storage to 30 days, adds clipping on visual objects, sound and emotion, text and timeline editing, and posting straight to YouTube, TikTok and Instagram. Pro is $29 per month.

Descript is the editor rather than the clipper. Transcript-based editing, filler word removal, Studio Sound and captions give you control over the result, from a free plan with 60 minutes of transcription per month and watermarked 720p exports, with Hobbyist at $16 per user per month billed annually.

Many people use both: automatic clipping to find candidates, an editor to finish the ones worth publishing.

When you don't need a tool

If you are not producing long-form video or audio, there is nothing to clip, and the honest first step is making something worth cutting up.

If you publish one episode a month, cutting two clips by hand in whatever editor you already own costs less than a subscription and gives a better result.

And if your short clips are not getting watched, more of them will not fix it. That is a hook problem: the first second, the caption, and whether the clip makes sense to someone who did not hear the previous ten minutes. Automatic tools are indifferent to all three, which is why the clips that perform are usually the ones a person edited afterwards.