How Long Should a Clip Be for Shorts and Reels?

Platform length limits: the maximum is not the target

The maximum length is what the platform will accept, and it says nothing about how long viewers will stay. Those are two separate questions, and mixing them up is how people end up padding a 20-second moment to 58 seconds.

YouTube Shorts: up to 3 minutes, vertical or square

Yes, a 3-minute video can be a Short. According to the YouTube Help Center, Shorts uploaded after October 15, 2024 can run up to 3 minutes, as long as they're vertical or square. Older Shorts still carry the 60-second limit, and that's why you'll see conflicting answers online.

And no, a 2-minute video isn't "too short" for YouTube. It's a normal video, or a Short if it's vertical or square and under the cap. Length doesn't disqualify anything.

Instagram Reels, Facebook Reels and TikTok: check the current cap

Those caps have changed several times, so we won't hard-code a number you might copy a year from now. Check the Instagram Help Center, the Meta Business Suite upload screen for Facebook Reels, and the TikTok upload screen in the app. Thirty seconds of checking beats a rejected upload.

Shorter usually holds more viewers, but story sets the length

Shorter clips usually get higher completion rate, but the story decides the length. Adobe's creator survey, for example, most often cited under 30 seconds as the ideal length. Treat that as one data point, not a rule.

It's a decent default, though. If you have a 12-second moment, posting 12 seconds is correct. Stretching it won't help, and cutting a payoff to hit a round number won't either.

Why completion rate drops when a clip drags

Viewers leave in the dead spots: the slow lead-in, the repeated explanation, the extra 5 seconds after the payoff. Each one gives a thumb a reason to swipe. We've watched clips lose most of their audience in a single flat stretch while the same clip, cut 8 seconds shorter, held well.

Remember the hook too. If the first 3 seconds don't earn attention, length is irrelevant because nobody reaches the end.

Starting lengths by clip type

These are starting points, not guarantees. They answer "best video length for Reels" and "ideal TikTok video length" better than one number can, because the right length depends on what's in the clip.

Podcast and interview clips

A podcast clip needs one complete thought. If the speaker makes a point in 35 seconds, 35 seconds is the clip. Cut the "so, anyway" before it and the tangent after it.

Sports and talent moments

These are short by nature. Start a beat before the setup so the result feels earned, then end on the reaction: the crowd, the judges, the celebration. Anything after that is filler.

Story and voiceover shorts

Here the beats set the length. A story with a hook, a turn and a payoff usually fits in 30-60 seconds with AI voiceover. If you need 90 seconds, ask whether the middle beat is doing any work. On YouTube, longer stories can use the 3-minute allowance, but only if retention holds.

How to choose the cut point: end at the natural end of the moment

Start just before the setup and end right after the payoff or reaction. That's the whole rule, and it's what "how long is a short clip" really comes down to: as long as the moment, plus a breath on each side.

Remove dead air inside the clip as well. A half-second pause between sentences feels natural in a conversation and slow on a phone. Keep the first 3 seconds strong, and make sure the last frame is the payoff, not a fade-out.

Setting min and max length in an AI clipper

In an AI Clipper, set a duration range rather than one target. For most clip types, 20-60 seconds is a sensible first range. The clipper proposes candidates inside that window, and then you do the editorial part: pick the one whose ending lands on the payoff.

Tighten the range for sports and talent (10-30 seconds) and widen it for podcasts. WowReveal's AI Clipper works this way with long videos, finding the best moments and rendering them as 9:16 vertical clips with captions. The range only narrows the search, and your eye still makes the final cut.

Test lengths with completion rate, not guesses

Compare completion rate and average view duration across lengths, not total views. Views depend on distribution luck. Completion rate tells you whether the clip itself held people.

Length alone doesn't decide eligibility or pay either. Originality and rule compliance matter far more for Facebook content monetization and the YouTube Partner Program, and Meta and YouTube make those eligibility calls. For rough earnings estimates, use a calculator such as the free Facebook earnings calculator, and treat the result as an estimate.

A simple 3-version test on one moment

1. Cut the same moment three ways: tight (ends on the payoff), standard (adds a short lead-in), and long (adds context and a longer tail).
2. Post comparable clips or variants over a similar window, without reposting the identical file to the same page.
3. Check completion rate and average view duration after the same number of days.
4. Keep the winning range for that clip type and repeat the test every few weeks.

Do this for podcasts, sports and stories separately. You'll end up with your own table, which beats any generic stat.