How Long Should a YouTube Video Be? The Honest Answer for 2026

There are two different questions hiding inside "how long should a YouTube video be", and mixing them up is why most advice on this topic is useless. One is a technical question about limits: how long can a YouTube video be before the platform refuses the upload. The other is a strategy question: how long do your videos need to be for people to actually watch them.

The first has an exact answer. The second does not, and anyone giving you a precise number is guessing. Here is what is verifiable, followed by how to work out your own answer.

How long can a YouTube video be?

A YouTube video can be up to 12 hours long or 256 GB in file size, whichever limit you hit first. That is the ceiling for a verified account, straight from YouTube's own upload documentation.

If your account is not verified, the cap is 15 minutes. Verification takes a couple of minutes through the mobile app using a text message or an automated call, and it is a one time step. If you have ever wondered why your upload was rejected at the 16 minute mark, that is why.

YouTube also notes that these limits have changed over the years, so you will occasionally find older videos on the platform that run past 12 hours. You cannot upload one of those today.

How long can a YouTube Short be?

Shorts can run up to 3 minutes. YouTube raised this from the original 60 seconds, and the multi segment camera in the Shorts tools records up to the full 3 minutes. Uploads are capped at 1080p resolution.

Worth knowing if you plan around metrics: since March 2025, a Short counts a view the moment it starts playing or replays, with no minimum watch time. Shorts view counts and long form view counts are not measuring the same thing, so comparing them directly will mislead you.

So how long should a YouTube video be?

YouTube itself does not publish a recommended length. Not in its creator documentation, not in its monetization rules, nowhere. That silence is meaningful. If a specific duration reliably performed better, the company sitting on all the data would have every reason to tell you.

What the platform optimizes for is watch time and satisfaction, and both of those are outcomes of whether the video holds attention, not of how many minutes it runs. A 6 minute video that keeps people to the end beats a 14 minute video they abandon at 3 minutes, and the reverse is equally true.

In practice, your video should be exactly as long as the idea can hold. Every minute past that point costs you retention, and retention is the thing the algorithm actually reads.

The 8 minute rule, and what it really costs

The one number that genuinely matters is 8 minutes. Videos of 8 minutes or longer can run mid roll ads, the ones that play during the video rather than before or after it. YouTube lowered this threshold from 10 minutes back in 2020 and it has stayed at 8 since.

This is where a lot of creators go wrong. They pad a 5 minute idea to 8 minutes to unlock mid rolls, and then wonder why the video underperformed. Padding does not fail neutrally. It damages the video, because the filler lands in the middle where viewers are already deciding whether to leave.

Videos under 8 minutes still earn. Pre roll, post roll and overlay ads all run normally. You are giving up one ad slot, not monetization.

The honest test is whether your idea supports 8 minutes of substance. If it does, take the mid rolls. If it does not, publish the 5 minute version and make the next video a bigger idea.

How long should your videos be when you are starting out

Shorter than you think, for a reason that has nothing to do with the algorithm. A new channel has no track record, so nobody is giving you the benefit of the doubt on a 20 minute runtime. More importantly, you are still learning to hold attention, and a shorter video gives you fewer minutes to lose people in.

There is also a compounding argument. Ten tight 6 minute videos teach you more about what your audience responds to than three sprawling 20 minute ones, because you get three times as many results to read.

Length changes by format, not by rule

A tutorial ends when the thing is built. A review ends when the verdict is delivered and defended. A video essay ends when the argument closes. A vlog ends when the day does. These are genuinely different shapes, and forcing them all toward one number is how you get bloated tutorials and rushed essays.

Ask what the format owes the viewer, then measure whether you have delivered it. That question produces a better runtime than any benchmark.

• Tutorial: as long as the steps take, with the setup cut to the bone
• Review or comparison: front load the verdict, then justify it
• Video essay or explainer: as long as the argument needs, which is usually longer than you are comfortable with
• Vlog: cut to the moments that actually happened, not the transitions between them
• Shorts: one idea, no setup, out before the loop restarts

When longer genuinely wins

There is a real case for length. A video that becomes the definitive answer to its question can run 25 minutes and outperform every 6 minute competitor, because it collects the searches, the rewatches and the "just watch this" recommendations for years afterwards.

But definitive is a quality bar, not a duration. You reach it by covering the failure cases and the edge cases everyone else skips, not by talking slower. If you cannot tell whether an idea wants to be 7 minutes or 20, script it first. A script makes a sagging middle visible before you film it, which is far cheaper than finding it in the edit.

Starting lengths by niche, if you have no data yet

None of these are rules. They are the ranges where established channels in each niche tend to settle once retention data has shaped them, which makes them a sensible first guess for a channel with nothing to measure yet. Treat them as a starting point you replace with your own retention graph after ten uploads.

• Tutorials and how-to: 8 to 15 minutes. Long enough to actually solve the problem, short enough that the viewer who came for one step does not leave at 40 percent.
• Commentary, news and reactions: 8 to 12 minutes. The topic, not the format, sets the ceiling, and padding a thin story to reach mid-roll length is the most common mistake here.
• Vlogs and day-in-the-life: 10 to 18 minutes. These live or die on pacing, so cut for rhythm and let the length fall out of the edit.
• Gaming and let's plays: 15 to 30 minutes, with much longer working once an audience is committed. Retention on gaming is driven by the game moment, not the clock.
• Product reviews: 8 to 12 minutes, plus a 60 second Short of the verdict. Reviews are the format where a companion Short most reliably feeds the long video.
• Documentary and video essays: 20 minutes and up, but only when the research is there. This is the one niche where longer genuinely wins, and the one where padding is most obvious.

How to find the right length for your channel

Stop guessing and read your own retention graph. In YouTube Studio, audience retention shows you an intro figure, which is the percentage of viewers still watching after the first 30 seconds, plus the spikes where people rewatched and the dips where they skipped or left.

Look at the dips first. If they cluster in the middle of your longer videos and not your shorter ones, your ceiling is lower than what you are publishing. If your long videos hold as well as your short ones, you have room to go longer.

This takes about ten minutes per video and it beats every generic benchmark on the internet, because it is measured on your audience rather than someone else's.

The short version

• A YouTube video can be up to 12 hours or 256 GB, whichever comes first
• Unverified accounts are capped at 15 minutes, and verification is a one time step
• Shorts can run up to 3 minutes at 1080p
• 8 minutes unlocks mid roll ads, and padding to reach it usually costs more than it earns
• YouTube publishes no ideal length, so use your own retention graph instead of a benchmark

Sources and methodology

We prioritize YouTube's own documentation for platform behavior and use third-party datasets only as broad benchmarks. Every external study has a different sample, geography, metric, and collection window, so its clock times should be tested against your channel rather than treated as a rule.

1. 01YouTube Help: Upload videos longer than 15 minutesThe 12 hour and 256 GB upload limits, and the 15 minute cap for unverified accounts.
2. 02YouTube Help: Manage mid-roll ad breaks in long videosThe 8 minute minimum for mid-roll ads.
3. 03YouTube Help: Get started creating YouTube ShortsThe current maximum length for a Short.

Frequently asked questions