The Perfect Instagram Reel Length for Maximum Impact

Want the exact frame size for every platform while you plan the cut? I keep a free cheat sheet updated: moonb.io/tools/social-media-sizes. No login, and it runs entirely in your browser.

The question people type into Google is “how long can an Instagram Reel be.” The question they actually mean is “how long should mine be,” and those have almost nothing to do with each other. I work at Moonb, and the team I sit with cuts client Reels most weeks of the year, so I have watched both versions of this play out. The max length is a technical fact you can look up in ten seconds. The right length is a decision, and it is downstream of one thing: whether people watch your Reel to the end.

Here is the rule I actually use on calls. Make it the shortest length that fully lands the idea, then let the retention data move you up or down. Everything else in this piece is me explaining why that rule beats any cheat-sheet number, and how to run it.

If you only want the numbers:

• The 2026 sweet spot is roughly 45 to 60 seconds. In a large 2026 analysis of business Reels, that band pulled both the highest engagement and the highest median views, close to double the views of clips under 30 seconds or over 3 minutes.
• Stay under 3 minutes to reach new people. Anything longer gets shown mostly to your existing followers, not to Explore or the Reels tab.
• Retention decides ranking, not duration. A 20-second Reel watched to the end beats a 90-second one abandoned at second 12, every time.
• Pick the shortest length that fully lands the idea, then test one shorter and one longer version of your best-performing format.

The Reel length sweet spot in 2026 is 45 to 60 seconds

If I had to hand you one band and walk away, it would be 45 to 60 seconds. In Socialinsider’s 2026 analysis of roughly 140,000 Reels from business accounts, the 45-to-60-second bracket posted the highest engagement rate of any length group and the highest median views, around 10,374, roughly twice what Reels under 30 seconds or over three minutes managed. Marketers land in the same place from a different direction: HubSpot’s video research found a third of them name 31 to 60 seconds as the optimal short-form length, and a majority say short-form video returns more than any other format they make.

So why did I open by telling you length is a decision, not a number, if the data points this cleanly? Because that band is an average across every account and every goal at once. It is a great starting bet. It is a terrible law. A trend-audio clip that needs 12 seconds will die if you pad it to 50 to hit the “sweet spot,” and a useful tutorial that needs 80 seconds will die if you guillotine it at 60. The average is where you start. The idea is what sets the actual length, and retention is what confirms it.

What actually changed: the 90-second maximum is a myth now

The single most repeated fact about Reel length is wrong, so let me clear it first. You are not capped at 90 seconds. In January 2025, Instagram head Adam Mosseri announced support for Reels up to three minutes, while adding that the platform is “still focused on short form video over long form.” By 2026 the technical ceiling stretched further, with uploads accepted up to 15 minutes and a 20-minute option rolling out to some creator and business accounts. The old 90-second wall people still quote came down years ago.

Here is the line that matters more than any of those ceilings. Reels over three minutes are not recommended to people who do not already follow you. They stay inside your existing follower graph instead of getting pushed out through Explore and the Reels tab. So the moment your Reel crosses three minutes, you have opted out of the discovery machine that makes Reels worth posting in the first place. In practice the real reach window is even tighter than that, which is why almost everything I have said sits under 90 seconds.

This creator walkthrough from Anita Wong is a clean screen-recorded demo of the expanded limits inside the app, if you want to see where the settings actually live:

For the full record, here is how the limit has climbed since Reels launched, and what each jump was actually about.

When	Max length	What the jump was actually about
Aug 2020	15 sec	Launch. A direct answer to TikTok's short, looping clips.
2021 to 2022	30, then 60 sec	Room for basic teaching and light storytelling beyond dances.
Feb 2023	90 sec	The number most guides still quote as the ceiling. It isn't anymore.
Jan 2025	3 min	Longer stories allowed, but this is also the reach cliff for non-followers.
2026	Up to 15 to 20 min (upload)	A technical ceiling, not a recommendation. Discovery still lives under 3 min.

Read that table as permission, not instruction. The fact that you can upload 15 minutes does not mean Instagram will show 15 minutes to anyone new.

Why the algorithm rewards the shortest complete version, not the longest

The ranking system does not ask how long your Reel is. It asks whether people finish it, rewatch it, and send it to someone. Mosseri has said plainly and repeatedly that a video’s fate is mostly decided in the first ten seconds or so, and that watch-through and completion are what the system optimizes for. Duration is not a lever the algorithm pulls. It is a side effect of the levers that matter.

Follow that logic and the whole “long versus short” debate collapses. A 90-second Reel that most people quit at second 12 is telling Instagram, in the only language it reads, that the content is not holding attention, so it gets shown to fewer people. A 25-second Reel with an 80 percent completion rate is telling it the opposite. Length only ever helps you by carrying an idea that keeps someone watching, and it only ever hurts you when it asks for time the idea has not earned. Short-form clips under a minute keep a completion-rate edge over anything past three minutes, and average active watch time on Reels sits somewhere around 19 seconds, which tells you how little runway you are actually working with.

This is why “cut, don’t pad” is the whole game. Every second that does not move the idea forward is a second where someone’s thumb gets bored and flicks up, and that flick is a vote against your reach.

The real sweet spot by goal: a retention-first cheat sheet

There is no universal perfect length, but there are sensible starting bands once you name what the Reel is for. I built this table around what the idea usually needs and what the retention data supports, not around invented KPI targets. Treat each band as your first draft length, then let completion rate edit it.

Notice the pattern down the middle column. As the goal gets more ambitious the band widens, but the ceiling on discovery-friendly content still sits at 90 seconds. If your instinct is to go past it, that is usually a sign the idea should be split into two Reels, not stretched into one. When I am mapping a client’s monthly slate, most of it lands in the 15-to-60-second range, and the longer pieces have to earn their length in the edit. For where these fit against the rest of your channel, our guide to short-form video Reels walks through the wider mix, and Stories video length covers the format people most often confuse with Reels.

How to find your ideal Reel length: cut, don’t pad

The method is boring and it works. Start from the idea, not the timer. Write or outline the Reel so it says exactly one thing, then time the shortest version that still says it fully. That number is your first cut, and it is almost always shorter than you expected. I have never once watched an edit and thought, this needed to be longer. The note is always the other way.

Then run the only test that settles it. Take a format that already performs for you, and make two versions of the same idea: one noticeably tighter, one with more room. Post them a week or two apart so you are not splitting your own audience on the same day. Compare completion rate and average watch time, not likes. The audience will tell you, in behaviour rather than opinion, where your ideal length actually sits, and it will differ by account and by topic. A/B testing turns “I think shorter is better” into “for us, 30 seconds beats 55 on tutorials,” which is the only version of this advice worth having.

One practical cut list I run on almost every rough edit, worst offenders first: drop the intro card, drop the “hey guys,” and drop the three seconds of breathing at the end after the point has landed. That alone tends to shave a fifth off the runtime without losing a single idea. When we build a client’s marketing video slate, this trimming pass is where most of the retention lift comes from, long before anyone touches fancier production.

You still win or lose in the first 3 seconds

Whatever length you land on, the first three seconds decide whether the rest of it gets watched at all. This is the mechanics of the feed, not a motivational line. Mosseri’s own framing puts the fate of a video in roughly its first ten seconds, and the drop-off inside a Reel is steepest right at the start, before anyone has committed. A great 60-second Reel with a weak opening performs worse than a fine 20-second one with a strong one, because the long version never gets the chance to be long.

A hook is a small promise: stay, and the next stretch of your attention will be worth it. Three formulas earn that promise more reliably than anything else I have tested with clients.

• The bold claim. Open on something surprising or slightly contrarian that challenges what the viewer assumes. “You are posting Reels at the wrong length” stops a thumb that “here are some Reel tips” never will.
• The provocative question. Ask something that lands on a real worry or want. “Is your best Reel getting buried because it is too long?” makes the viewer answer in their head, and answering means staying.
• The action-packed open. Skip the setup and drop straight into the most interesting moment, the “after,” the payoff, the result. The missing context becomes a small mystery the viewer sticks around to solve.

The mechanics underneath the hook matter as much as the words. Lead with movement rather than a static frame, put a bold readable text overlay up in the first second because most feed video is watched on mute, and make sure your opening line or sound is worth un-muting for. If you want the deeper playbook, our guide on how to create video hooks goes further than I can here.

How to measure it: the metrics that actually decide length

Views and likes will lie to you about length. To know whether your duration is right, open Instagram Insights on the Reel and read the four numbers that describe behaviour instead of applause.

• Completion rate. The share of viewers who reached the end. A high number on a short Reel is expected; a high number on a 75-second one is proof the length was earned.
• Average watch time. How far into the Reel people got, on average. When this climbs as you shorten a format, you have found a length that is too long. When it holds steady as you extend, you have room.
• Saves. A save means the content was worth keeping. These tend to run higher on the teaching-length Reels in the 45-to-60-second band, where there is enough substance to bookmark.
• Shares. A share is a public endorsement, and it skews toward the shortest, most rewatchable clips. If shares are your goal, the data is telling you to cut harder.

Read these together and length stops being a guess. Rising completion plus rising watch time as you trim is the signal to keep trimming. Falling watch time when you extend is the signal you overreached. That feedback loop, run a few times on your own account, is worth more than any benchmark I or anyone else can hand you, because it is measured on your audience rather than an average of everyone’s.

The one line I will add about doing this at volume: the method above is simple to run once and hard to run on every Reel, every week, without the quality sliding. Keeping the cut tight and the hook sharp across a full slate is the part my team owns for the brands we work with, so the length question gets answered by the edit instead of a spreadsheet. If that is the part eating your week, see how Moonb works.