LinkedIn video length: limits and best practice

LinkedIn allows a member to share native video between 3 seconds and 15 minutes long from desktop, with a 2 second minimum from the mobile app and a file size between 75KB and 5GB, according to LinkedIn's own video sharing help page. Video posted by a LinkedIn Page is documented separately and capped at 10 minutes, which is where the widely repeated “LinkedIn videos can be 10 minutes” figure comes from. In practice you should be aiming at 30 to 90 seconds, and the length is rarely what decides whether a video works. Most video posts fail because they are silent to the majority of the people who see them, since video autoplays muted and uncaptioned video is unreadable the moment sound is off.

What is the maximum LinkedIn video length?

Fifteen minutes for a member post, ten minutes for a Page post. Those are two different documented limits and confusing them is the reason this question gets answered inconsistently across the web.

Sources for both columns are LinkedIn's own documentation: video sharing troubleshooting for member posts and video specifications for LinkedIn Pages and Career Pages for the Page column. These limits have changed over time, so if you are reading an older guide that gives a single number, check which surface it was describing and when it was written.

A note on file formats, which is where uploads actually fail. The member page lists MP4, MOV, AVI, WEBM, MKV, WMV, VC1, MPEG, MPEG2, MPEG1VIDEO, DVVIDEO, QTRLE and TSCC2 as supported on desktop, iOS and Android. MP4 with H.264 is the safe default and the one least likely to produce a processing error twenty minutes after you hit post.

How long should a LinkedIn video actually be?

Thirty to ninety seconds for most posts. We want to be clear about the status of that recommendation before you act on it: it is judgment about how feed attention works, not a measured finding, and we will explain in a moment exactly what our data can and cannot tell you about duration.

Length is a budget you spend against attention. Every additional thirty seconds has to earn its place.

The mistake that costs the most is not total runtime. It is the first fifteen seconds. “Hi everyone, hope you are having a great week, I wanted to jump on here and talk about something” spends the entire attention budget on nothing, and it appears in a remarkable share of professional video. Cut it and a ninety-second video becomes a seventy-five second video that works.

Which lengths suit which kinds of video?

The right runtime depends less on a platform average than on what you are filming. Five formats cover most of what professionals actually make.

• Screen recording with voiceover: 60 to 120 seconds. The single most useful video most professionals can make, because you are showing something only you have access to. It needs longer than a talking head because the viewer is reading the screen as well as listening.
• Straight-to-camera opinion: 30 to 60 seconds. One argument, one example. Past a minute of an unbroken talking head, completion falls away sharply unless the content is unusually good.
• Product or process demo: 90 seconds to 3 minutes. Only if something is visibly happening throughout. A demo where the interesting part arrives at 2:10 should start at 2:05.
• Event or behind-the-scenes clip: 15 to 45 seconds. These work on atmosphere rather than argument, and atmosphere does not sustain three minutes.
• Teaching something genuinely complex: 3 to 10 minutes, rarely. Only for an audience that already knows you. This is really a newsletter or a hosted video that happens to be uploaded to the feed.

A useful test across all five: could you cut the video in half and lose only detail, not the point? If yes, cut it in half. The half you keep will outperform the whole.

What does our data say about LinkedIn video?

It says a lot about whether to use video and nothing about how long to make it, and the second half of that sentence matters.

In our analysis of 12,988 posts, native video made up 26.3% of the top decile against 10.1% of the bottom half, a 2.6x over-index across 1,613 video posts from 46 creators. The median video post in the cohort earned 181 reactions and 27 comments at a rate of 0.60 per 1,000 followers, against a cohort median of 122 reactions, 12 comments and 0.40. The full format breakdown, including how video compares with images, documents and link shares, is in our analysis of LinkedIn video posts.

What the dataset does not contain is the duration of any of those 1,613 videos. It records that a post was video, not how many seconds it ran. So we cannot tell you that forty-five seconds beats three minutes, and neither can anyone else working from engagement data alone. Treat any page giving you a precise optimal length with a percentage attached as unsourced until it shows you the measurement.

Where the segment data is useful is in showing how strongly video separates good posts from weak ones once you hold the subject matter constant.

Read those as correlations with a selection problem baked in. Nobody films, edits and uploads a video casually, so part of what these gaps measure is effort rather than format. The video sample is large enough that effort cannot be the whole story, but it is certainly part of it.

Why do most LinkedIn videos fail on sound rather than length?

Because video autoplays muted in the feed. A viewer scrolling past your video is watching a silent film unless you put the words on the screen, and no amount of tightening the runtime fixes a video whose entire content is spoken.

The most cited figure here is that up to 85 percent of social media videos are played with no sound, reported by Hootsuite in its guide to LinkedIn video, attributed to LinkedIn. That page carries a last-updated date of April 2021, so we would treat the specific number as old. The direction is not seriously disputed by anyone, and you can confirm it on yourself: count how many videos you unmuted this week.

The accessibility case is stronger than the reach case and gets mentioned less. An uncaptioned video excludes deaf and hard-of-hearing viewers entirely. It also excludes anyone in an open-plan office, on public transport, next to a sleeping child, or reading in their second language. Publishing without captions is a decision about who is allowed to understand you, which is the same category of decision as using Unicode characters to fake bold text, and it is just as avoidable.

How do you add captions to a LinkedIn video?

Two routes, and they are documented separately by LinkedIn.

Auto captions. LinkedIn's auto captions help page states that auto captions are available for all LinkedIn members and Pages, and that they are currently supported in English, Spanish, Portuguese, French, German, Italian, Indonesian, Turkish, Arabic and Dutch. There is a toggle in the editor to add them automatically, and a separate toggle to review them before viewers can see them. LinkedIn also notes that captions for supported languages may not always be available and that they can take time to appear after posting. For scheduled posts, captions only become reviewable once the video is live.

Your own SRT file. LinkedIn's closed captions documentation states that you must have an associated SRT file attached to the video before it can be posted, using the Select Caption control during upload, and that closed captions can be added by personal accounts or Pages.

Practical guidance on top of the mechanics:

• Turn the review toggle on the first few times. Automatic transcription is reliable until it hits a product name, an acronym, or an accent, and then it is confidently wrong in a way that makes you look careless.
• Consider burning captions in as well. Burned-in text always displays, survives being downloaded or reshared, and does not depend on a viewer toggling anything. The trade is that it is not machine readable, which is why an SRT or auto caption track is still worth having alongside it.
• Keep the on-screen text large and high-contrast.Most viewing happens on a phone held at arm's length, often outdoors.

What should the first three seconds do?

Exactly what the first line of a text post does: give someone a reason not to keep scrolling. The structural logic is identical, which is why the advice in how to write LinkedIn hooks transfers almost without modification.

• Open on the claim, the question, or mid-scene. Never on a title card or an intro animation.
• Have your face or the subject already in frame and already moving.
• Put the point in the on-screen text within the first two seconds, for muted viewers.
• Write a caption above the player that works as a post on its own, because it is doing the stopping.

The caption above the video is not a summary of the video. It is the reason someone watches at all. In our cohort, video captions say you somewhere in the visible hook 43.0% of the time against 36.4% cohort-wide, and open with a question 16.6% of the time against 11.4%. Video creators ask more questions, and questions correlate with comments.

Does aspect ratio matter more than length?

For completion rate, usually yes. LinkedIn accepts anything between 1:2.4 and 2.4:1, and within that range the practical question is how much of a phone screen your video occupies.

A 16:9 landscape video on a vertical feed is a letterboxed strip in the middle of the screen, competing with everything above and below it. A square or vertical video fills the frame. That is a larger difference to whether someone watches than fifteen seconds of runtime, and it costs nothing to fix at export time.

Square is the safe default if you also use the video elsewhere. Vertical is stronger on LinkedIn specifically and weaker everywhere a desktop viewer is likely. If you film once and post in several places, film with enough headroom to crop.

How do you tell whether your video was too long?

Check the numbers before you change the format, because most people who conclude “my videos are too long” actually have an opening problem.

LinkedIn's post analytics documentation lists video views, watch time and average watch time among the metrics available, along with impressions, members reached, followers gained from the post, and viewer demographics. It also states that video analytics are available for 365 days and that the numbers are estimates rather than precise counts.

• If average watch time collapses in the first few seconds, the opening is the problem. Cut the preamble and refilm the first ten seconds only.
• If it holds and then drops in the middle, you have a structure problem. Something in the middle stops paying off, usually a section that could be one sentence.
• If it stays high to the end and the view count is low, the video is working and the caption or thumbnail is not. Do not shorten a video people are finishing.
• If views are high and comments are near zero, the ending is not asking for anything. That is a call to action problem, not a length one.

One warning about comparing numbers across formats. Video views are counted on a different basis from impressions, so a video “view” and a text post impression are not the same unit and should not be put in the same chart. We untangle that in impressions versus reach.

Do longer LinkedIn videos get less reach?

There is no evidence we would stand behind either way, and this is a question where the confident answers online are all built on the same unsourced assumption.

The assumption runs like this: LinkedIn can observe how long people watch, watch time is a signal of interest, therefore videos that hold attention are distributed more widely, therefore shorter videos with higher completion rates win. Each step is plausible. LinkedIn has never published a statement confirming the chain, and our dataset cannot test it because it does not record duration. So the honest position is that a video which holds attention is probably better positioned than one that does not, and that this is a statement about attention rather than a rule about seconds.

What that means practically is that completion rate matters more than runtime, and the two are not the same thing. A three-minute video watched to the end is a better outcome than a forty-second video abandoned at eight seconds, even though the second one is shorter. Optimising for a target duration rather than for whether the content sustains itself is how people end up padding a strong forty-second point to ninety seconds because they read that ninety was optimal.

The one length-adjacent claim we are comfortable making from our own data is about format, not duration: native video appears far more often among top posts than among weak ones, and shared links run the opposite way. Uploading the file rather than linking to it elsewhere is the decision that moves the needle. We work through why in our analysis of external links and LinkedIn reach.

What do you do if the file is too big or the upload fails?

Uploads fail more often than they should, and almost always for reasons the spec table above predicts.

1. Check the duration against the right limit. If you are posting as a Page, your ceiling is 10 minutes rather than 15. A 12-minute video that works from a personal profile will not post from a Page.
2. Re-export as MP4 with H.264. Several exotic formats are listed as supported and MP4 is the one that consistently processes without incident.
3. Bring the bit rate inside 192 Kbps to 30 Mbps. Exports from professional editing tools regularly exceed 30 Mbps at high resolutions, which produces a needlessly large file for no visible benefit in a feed.
4. Keep the frame rate between 10 and 60 fps. Anything filmed at 120 fps for slow motion needs conforming before upload.
5. Stay well under 5GB. The limit is generous and a 1080p feed video has no reason to approach it. Smaller files upload and process faster, which matters when you are posting on a phone.
6. On iOS, stay in the app while it uploads. LinkedIn documents that iOS uploads can stall if you switch away, and that videos stored in iCloud rather than on the device cannot be uploaded from the mobile app.

When is a long LinkedIn video the wrong call?

Three cases where the runtime is genuinely the problem rather than a symptom.

Repurposed webinars and podcast clips. Those recordings were made for an audience that chose to be there and had already heard the introduction. A clip that starts mid-sentence in a conversation nobody heard the beginning of is the most common bad video on LinkedIn, and it is usually also the longest.

Anything dense or numeric. Nobody can hold a table in their head from video. If the content is a comparison, a set of figures, or a definition people will want to reread, put it in text or in a document post where the reader controls the pace. We make that case in the piece on LinkedIn carousels.

Anything time-sensitive. A reaction to news is worth more within the hour than it is polished. By the time you have filmed, captioned and uploaded, the conversation has moved on.

How do you pick a length in sixty seconds?

1. Say the point out loud once and time it. That is your floor. Most points take twenty to forty seconds.
2. Add an example, not an explanation. One concrete instance is worth more than a minute of elaboration.
3. Cut the first ten seconds of whatever you recorded. Almost always warm-up, almost always improves it.
4. Stop when the point has landed. Trailing off at the end is the second most common flaw after preamble at the start.
5. Caption it before you consider it finished. An uncaptioned video is not a shorter video, it is an incomplete one.

The short answer on LinkedIn video length

LinkedIn video length runs from 3 seconds to 15 minutes for a member post and up to 10 minutes for a Page post, with a 75KB to 5GB file size range, aspect ratios between 1:2.4 and 2.4:1 and resolutions up to 4096x2304, all documented by LinkedIn directly. The useful range for the feed is 30 to 90 seconds, offered as judgment rather than as a measured optimum, because our dataset records that a post was video without recording how long the video ran. What our data does support is that native video appears far more often among the best posts than the worst, at 26.3% against 10.1% across 1,613 videos from 46 accounts. Spend your effort on the first three seconds, the caption above the player, and captions inside it, and the exact runtime will mostly take care of itself. The full format comparison sits in our study of 34,000 LinkedIn posts.