How many Shorts can you cut from one long video?
How many Shorts can you get from one long video?
Divide the source length by the clip length. Most tools cut sequentially at roughly fifty-five seconds, which keeps every clip safely under the one-minute mark, so a ten-minute video yields about ten clips and a three-minute video yields three. Asking for more does not produce more, because the clips are consecutive slices of a finite timeline rather than variations on a theme. There is no second pass that finds extra moments once the footage runs out. This is why tools that let you type any number can appear broken: you request thirty, you receive three, and nothing explains the gap. If you want thirty clips you need roughly thirty minutes of source material, which in practice means a podcast episode, a long tutorial or a stream recording rather than a single talking-head take.
The arithmetic is simple and almost nobody states it, so people reasonably assume the tool is misbehaving when it isn’t.
The formula
clips = floor(source duration ÷ clip length)With a 55-second clip length, which is what most tools use to stay comfortably under the one-minute threshold:
| Source video | Clips you can get |
|---|---|
| 1 minute | 1 |
| 3 minutes | 3 |
| 5 minutes | 5 |
| 10 minutes | 10 |
| 30 minutes | 32 |
| 1 hour | 65 |
That’s the whole thing. A 192-second video contains three 55-second slices and a 27-second remainder. There is no fourth clip in it.
Why asking for more doesn’t work
The mental model people bring is something like “the AI watches the video and finds the best moments,” which implies that asking for more moments should surface more of them. If the tool worked that way — picking highlights from anywhere in the timeline, allowing overlap — then thirty clips from a three-minute video would be possible, just repetitive.
That is not how sequential clipping works. The video is cut into consecutive, non-overlapping slices: 0:00–0:55, 0:55–1:50, 1:50–2:45, and then the footage is gone. Requesting thirty from three minutes of source is arithmetically the same as requesting thirty pages from a three-page book.
What the AI is actually doing is writing the metadata for each slice — the title, the hook, the description and tags — not deciding where the slices fall.
The 55-second choice
Why 55 and not 60? Because platform limits are edges you don’t want to sit on.
A clip that renders at 60.2 seconds because of frame rounding is no longer a Short on some paths, and the failure shows up after the render rather than before. Cutting at 55 leaves five seconds of headroom for encoding variation, and viewers do not perceive the difference.
It also means a 60-second source gives you exactly one clip, not one-and-a-bit, which is usually what people want anyway.
What this means for choosing source material
If you want volume, the source length is the only lever that matters. Working backwards from a posting schedule:
- One Short a day for a month = 30 clips ≈ 28 minutes of source video.
- Three a week for a month = 12 clips ≈ 11 minutes.
- One a week for a quarter = 13 clips ≈ 12 minutes.
A single ten-minute tutorial is a week and a half of daily posting. A 45-minute podcast episode is a month and a half. This is why long-form creators get so much more out of clipping tools than people recording bespoke short videos — the maths is entirely on their side.
Quality falls off before the footage does
One honest caveat against just uploading the longest thing you have: not every 55 seconds of a 45-minute recording is worth posting. Sequential clipping will happily hand you the segment where someone was adjusting a microphone.
In practice the useful yield is lower than the arithmetic maximum — for most long recordings, somewhere between half and two-thirds of the slices are worth publishing. The right workflow is to generate the full set, watch them at speed, and throw away the ones that don’t open well. That is much faster than finding the good moments by hand, which is the actual time saving.
So the realistic version of the table above is closer to:
| Source video | Clips generated | Clips worth posting |
|---|---|---|
| 10 minutes | 10 | 5–7 |
| 30 minutes | 32 | 16–22 |
| 1 hour | 65 | 30–45 |
Still a very good trade. Just not a machine that turns three minutes into a month of content, and any tool implying otherwise is selling you something.
If you only have short source material
Two options, neither of them a trick:
Record longer. The single highest-leverage change. A 20-minute unedited talk-through of something you know produces twenty clips; twenty separate 1-minute recordings produce twenty clips and take four times as long to make.
Post fewer, better. Three strong Shorts a week beats fifteen weak ones, and the algorithm agrees with that more than most people expect.
Bulk Scheduler queues YouTube Shorts, long-form videos and Instagram Reels from one library. The Free Demo is one video, five AI Shorts and five scheduled posts — no card, no countdown.