Eight minutes, for mid-roll ad eligibility. That's it. Everything else is a pattern in the data, not a setting in the system — here's what the pattern actually says.
Verified against YouTube Help and 2026 reporting, August 2026 · Re-verify before publishing
Every "ideal length" you've read is a pattern. Only one number is a rule.
8 minutes, for mid-roll ad eligibility. Nothing else is enforced.
Every creator eventually runs into the same conflict. The advice says longer videos earn more. The analytics say viewers are leaving at ninety seconds. So which do you listen to?
The honest answer is that this is a false choice, and the reason it feels like one is that most length advice is written as if YouTube enforces a rule it does not actually enforce. There is exactly one length threshold YouTube genuinely applies: eight minutes, for mid-roll ad eligibility. Everything else you've read about ideal duration is a pattern observed in data, not a setting in the system.
| Format | Working range |
|---|---|
| General long-form | 7 to 15 minutes covers most content and most channels |
| The monetization line | 8 minutes — the only hard threshold |
| Shorts | Up to 3 minutes qualifies. The data favors the longer end. |
| The full span | A 15-second clip and a three-hour deep dive can both work. Neither is wrong by length alone. |
A video should be as long as it takes to deliver its value, and not one second longer.
A tightly edited 9-minute video and a bloated 9-minute video are the same length and completely different products. Length describes the container, not the content.
Which is why the sections below give ranges rather than targets, and why Part 6 — finding your own number from your own analytics — is the one that actually matters.
If you take one thing from this page: never add time. Only ever remove it, until removing more would cost the viewer something.
| Content type | Working range | Why |
|---|---|---|
| Tutorials and how-tos | 8–15 minutes | Long enough to break down a process cleanly without skipping troubleshooting — usually the part viewers came for. Also lands above the mid-roll line naturally. |
| Product reviews and unboxings | 5–12 minutes | Enough to show real use, specs, and pros/cons. Below 5 minutes tends to read as promotional rather than evaluative. |
| Vlogs and storytelling | 10–20 minutes | Carried by personality and narrative momentum rather than information density. |
| Explainers and educational | 8–15 minutes | Needs room for a mechanism, not just a conclusion. |
| Commentary and reaction | 10–20 minutes | Note the reused content policy — substantial original commentary is required regardless of length. See our faceless channel guide. |
| Podcasts and interviews | 30 min to 2+ hours | Audience arrives expecting duration. Chapters matter enormously here. |
| Shorts | Up to 3 minutes | See the correction below — most advice on this is out of date. |
YouTube Shorts now accepts vertical video up to three minutes. The 60-second ceiling most advice still repeats is gone.
And the largest platform dataset available disagrees with "shorter is better." Across 756,851 videos analyzed by RevID in 2026: 0–15 seconds was 4.4%. 16–30 seconds was 11.9%. 31–60 seconds was 31.1%. 61–90 seconds was 16.9%. And 90+ seconds was 35.7% — the single largest bucket.
The average exported short-form video ran 115.5 seconds — just under two minutes.
| Short-form duration | Share of videos | Read |
|---|---|---|
| 0–15 seconds | 4.4% | Ultra-short hooks, teasers, memes |
| 16–30 seconds | 11.9% | The classic short-form range — now a minority |
| 31–60 seconds | 31.1% | Sweet spot for educational content |
| 61–90 seconds | 16.9% | Extended storytelling |
| 90+ seconds | 35.7% | Largest bucket — long-form shorts, explainers |
| Sub-30 total | 16.3% | — |
Source: RevID.ai, "The Anatomy of a Viral Short," analysis of 756,851 videos, March 2026. Platform-specific to RevID and to AI-generated video — the report says so itself. Directionally useful, not universal.
One creator-tools analysis cites a study finding Shorts around 50 seconds outperformed all other lengths.
That doesn't contradict the data above so much as sit inside it — 50 seconds falls in the 31–60 second band, the second-largest bucket at 31.1%.
The defensible synthesis: the useful Shorts range is roughly 30 to 90 seconds, and the instinct to cut everything to 20 seconds is costing people watch time.
The algorithm does not rank duration. It ranks what duration produces.
| What YouTube measures | What it does not measure |
|---|---|
| Total watch time | Video length as a standalone factor |
| Audience retention — percentage viewed | Whether you hit an arbitrary minute mark |
| Average view duration | — |
| Click-through rate on impressions | — |
| Session behavior — what viewers do next | — |
It's a balance rather than a rule. Videos in this bracket accumulate meaningful watch time per view while keeping the drop-off curve manageable. A 25-minute video has a higher ceiling and a much steeper cliff; a 4-minute video has a gentle curve and a low ceiling.
The tightly edited 10-minute video beats the bloated 25-minute one on total watch time despite being 60% shorter, because retention is doing more work than duration.
It also beats the 4-minute video, because at some point high retention on a short video runs out of room.
The optimum isn't the longest video you can make or the shortest. It's the longest video you can make without the retention curve collapsing — and that number is specific to your channel.
| Scenario | Length | Retention | Watch time | Outcome |
|---|---|---|---|---|
| Tight edit | 10 min | 60% | 6.0 min | Wins |
| Padded | 25 min | 20% | 5.0 min | Loses despite 2.5× the length |
| Very short | 4 min | 80% | 3.2 min | Great retention, low ceiling |
| Padded to threshold | 8 min | 25% | 2.0 min | Worst of both — see Part 4 |
That last row is the one to watch. A 4-minute idea stretched to 8 minutes to unlock mid-rolls doesn't just fail to help — it produces less watch time than the honest 4-minute version would have.
You've traded away distribution to gain an ad slot on a video the algorithm has stopped recommending. The mid-roll is worth having. It is not worth having at the cost of the retention curve.
| Rule | Detail |
|---|---|
| Minimum for mid-rolls | 8 minutes of total video runtime |
| History | Lowered from 10 minutes to 8 minutes in 2020. Unchanged as of 2026. |
| What counts | Final uploaded duration — intro, outro and bumpers included. A 7-minute topic that runs 8:30 with bumpers qualifies. |
| Under 8 minutes | Still eligible for pre-roll and post-roll ads. |
| Shorts | Not eligible for mid-rolls — Shorts monetize through the Creator Pool instead. See our Shorts monetization guide. |
| Back catalog | Any monetized long-form video in the YPP qualifies, including older uploads. |
Sources: YouTube Help, "Manage mid-roll ad breaks in long videos"; vidIQ, "YouTube Mid-Roll Ads: 8-Minute Minimum & Placement Rules," May 2026.
YouTube will automatically show mid-roll ads at natural break points — like transitions and pauses — starting May 12, stepping away from reliance on manual ad placements. — eMarketer, reporting on YouTube's announcement
| Change | What it means for you |
|---|---|
| Automatic placement at natural breaks | YouTube identifies pauses, scene changes and transitions rather than inserting at fixed intervals. |
| Older videos updated | YouTube applied automatic ad slots to back-catalog videos. |
| Studio feedback feature | YouTube Studio now flags manual ad slots it considers interruptive, giving you a chance to move them. |
| Feedback timing | Initial feedback generally within an hour of upload. Adjustments after that report back in under a minute. |
| Manual plus automatic | You can run both — YouTube keeps your manual breaks and adds slots at natural points. |
Place your manual mid-rolls, upload, wait an hour, and YouTube will tell you which ones it thinks are interruptive — then move them and get an answer in under a minute.
That turns ad placement from guesswork into a short feedback loop, and it's available to anyone in the YPP right now.
| Video length | Suggested mid-rolls | Placement |
|---|---|---|
| 8–12 minutes | 1 | At the most natural structural break after roughly the 5-minute mark |
| 12–20 minutes | 2 | At natural transitions, not within 3 minutes of each other |
| 20–40 minutes | 3–4 | At genuine content transitions |
| General | — | Avoid the first and last two minutes — lower completion rates |
Source: TubeMilestone, July 2026, which presents these explicitly as "starting points, not rules." Treat them that way.
One monetization guide puts it plainly: do not artificially pad videos to reach 8 minutes — thin content reduces watch time and hurts overall channel performance more than mid-rolls help.
The legitimate version: if a topic naturally supports more depth — an extra worked example, a troubleshooting section, a second use case — adding it to cross 8 minutes is fine. If you're stretching, stop.
Several sources cite figures like "$1–3 RPM pre-roll only versus $2–6+ with mid-rolls," or a "+30–40% RPM lift." Those come from AI video tool marketing sites and are not substantiated. RPM varies enormously by niche, geography and season — see our niche-by-niche real data in the faceless channel guide.
Better than any figure you'll find online: compare your own under-8 and over-8 videos in the Studio Revenue tab. It's free and specific to you.
Every article on this keyword gives durations. Almost none gives the conversion, which is what a creator actually needs at the moment they're planning a video.
| Target runtime | Words at 150 wpm | Note |
|---|---|---|
| 60 seconds | 150 | Short |
| 90 seconds | 225 | The largest short-form bucket |
| 3 minutes | 450 | Short-form ceiling |
| 5 minutes | 750 | Under the mid-roll line |
| 8 minutes | 1,200 | The mid-roll threshold |
| 10 minutes | 1,500 | — |
| 12 minutes | 1,800 | — |
| 15 minutes | 2,250 | Upper end of the general sweet spot |
| 20 minutes | 3,000 | — |
Our own measured pacing, from a real render: an 880-word test script rendered at exactly 352 seconds — 150 words per minute, on the nose. That's the figure used throughout this table and our script-writing guide.
Word-count estimates from a script draft are a planning tool, not a guarantee. Voice, pacing settings, and pauses all shift the real number. If you're targeting 8 minutes specifically for the mid-roll threshold, confirm the finished render actually crosses 8:00 before you publish — don't assume from the word count alone.
No, not as an algorithm rule — and it never really was one. The belief that videos had to hit 10 minutes to earn algorithmic favor came from the old mid-roll threshold, which YouTube lowered to 8 minutes in 2020. People remembered the number and forgot the reason.
What remains true is narrower and purely operational: crossing 8 minutes unlocks mid-roll ads, and mid-rolls are a meaningful share of long-form ad revenue. That's a monetization decision, not a ranking one. If your video is genuinely a 5-minute video, publish a 5-minute video. You lose the mid-roll and keep the retention curve — and the retention curve is what gets you recommended.
Only if the content genuinely breaks in two — a natural split, a clear before-and-after, separate projects. An artificial split doesn't work: cutting a single argument in half produces a Part 1 that never pays off and a Part 2 most people never find. The better tool in most cases is chapters, which let viewers navigate without leaving the video and preserve the watch time a split would have scattered.
Stop reading benchmarks and open your Audience Retention report. Every number on this page is an average of other people's channels; your own analytics describe yours.
Look at where the curve drops. A steep cliff in the first 30 seconds is a hook problem, not a length problem — see our script-writing guide for the hook data. A steady decline that falls off a shelf at a consistent point tells you where your audience runs out of patience — and that point is your ceiling.
Then compare your videos against each other. Sort by average percentage viewed rather than views, find your three strongest and three weakest, and look at what they have in common. If your best performers cluster around 6 minutes, your audience has told you something more useful than any industry benchmark could.
Real renders, real pacing, real numbers — 20% off RevID.ai with code THEAIBUILDLAB.
Try RevID.ai →| Question | Answer |
|---|---|
| What is the maximum YouTube video length? | 12 hours for verified accounts. Rarely the binding constraint. |
| How long can a YouTube Short be? | Up to 3 minutes — not the commonly repeated 60 seconds. |
| Do longer videos rank better? | No. Watch time and retention rank; length only influences them. |
| Does length affect Shorts monetization? | Shorts monetize through the Creator Pool, not mid-rolls. See our Shorts monetization guide. |
| How many words is a 10-minute video? | About 1,500 at 150 words per minute. See Part 5. |
| Should my first videos be short? | Generally yes — easier to keep tight while you're learning to edit. |
| Do chapters help long videos? | Yes, for navigation and retention. They don't fix pacing. |
| Mistake | Consequence |
|---|---|
| Padding to reach 8 minutes | Retention collapses and you lose more in distribution than you gain in ad slots. |
| Believing 10 minutes is an algorithm rule | It's not, and it never was. It was the old mid-roll threshold. |
| Cutting Shorts to 20 seconds by default | 90+ seconds is the largest bucket at 35.7%. Sub-30 is 16.3%. |
| Saying Shorts max out at 60 seconds | They accept up to 3 minutes. |
| Chasing a benchmark instead of reading retention | Your own analytics beat every average on this page. |
| Splitting a video that doesn't break naturally | Part 1 never pays off; Part 2 never gets found. |
| Judging length by views instead of average percentage viewed | Views measure the thumbnail. Retention measures the video. |
| Ignoring the Studio ad-placement feedback | It tells you which mid-rolls are interruptive, within an hour. |
| Planning a mid-roll strategy on estimated runtime | Confirm the finished file crosses 8:00. |
| Trusting RPM figures from tool marketing sites | Compare your own under-8 and over-8 videos in the Revenue tab. |
| Adding time to any video, ever | Only ever remove. |
The 8-minute mid-roll rule is confirmed current across multiple 2026 sources and consistent with YouTube Help — this is the load-bearing claim on this page, and it's worth re-verifying before you publish anything that states it as settled. The RevID duration data is platform-specific and AI-video specific by the report's own admission. Content-type length ranges in Part 2 are conventional industry guidance, not measured data — treat them as working ranges. Mid-roll placement guidance in Part 4 is presented by its own source as a starting point, not a rule. RPM figures circulating online for "pre-roll vs. mid-roll" comparisons come from AI tool marketing sites and are not substantiated — we deliberately don't repeat them here. One minor source conflict: a YouTube Help snippet references mid-rolls "on Shorts or videos longer than 8 minutes," while vidIQ states Shorts are not eligible; the Shorts-not-eligible reading is the clearer and more consistent one, but worth confirming directly if it matters to your plan.