How much space a downloaded video actually takes
A ten minute video at 720p came to 161 MB when we measured it on our own server, which is about 15 MB per minute. The same video at 360p came to 28.8 MB, and as an MP3 it came to 19.0 MB. Those numbers are one measured example rather than a rule, because the thing that really drives file size is bitrate, and bitrate moves with frame rate and with how much is happening on screen. This guide gives you the measurements, explains what makes them move, and shows how to reason about a file before you commit to downloading it.
The measured example
We downloaded Big Buck Bunny, 10 minutes 35 seconds from a 60fps source, on our own server and weighed the results. Almost nobody publishes numbers for this, and the one page that does gives a single figure with no per minute breakdown, so here is the whole thing.
| Format | File size | Per minute |
|---|---|---|
| MP4 720p | 161 MB | about 15 MB |
| MP4 360p | 28.8 MB | about 2.7 MB |
| MP3 | 19.0 MB | about 1.8 MB |
Read this as one example, not a universal law. A 60fps animation sits at the heavy end of the range, and a static talking head at 30fps comes in well below these numbers at the same resolution.
Bitrate is what actually decides the size
File size is bitrate multiplied by duration, and resolution is only one of the things that sets bitrate. That is why two 1080p files of the same length can differ by a factor of three, and why the resolution label on its own is a weak predictor.
Bitrate is the number of bits used per second of video. An encoder spends more of them when the picture is hard to compress and fewer when it is easy, so the same resolution can be encoded thin or generously depending on what the platform decided the content needed.
The practical consequence is that you should treat any per minute figure, including ours, as an anchor rather than a promise. Doubling the resolution does not simply double the size, and halving it does not simply halve it.
- Frame rate: 60fps carries twice as many pictures per second as 30fps, and costs accordingly.
- Motion: fast action, camera movement, confetti, rain and grain all resist compression and push bitrate up.
- Simplicity: a static shot with a plain background compresses down to very little.
- Duration: the one factor that scales exactly, which is why a per minute figure is the useful unit.
An upper bound you can reason with
YouTube publishes recommended upload bitrates, and they make a useful ceiling. Creators are told to hand in 8 to 12 Mbit for 1080p, 35 to 68 Mbit for 2160p30 and 53 to 85 Mbit for 2160p60.
Those are upload recommendations, and the streams YouTube serves back are lower than what it asks creators to submit, because it transcodes the upload for delivery. So treat these figures as an upper bound on what a download could weigh, never as an estimate of what it will weigh.
What they do tell you clearly is the shape of the curve. Going from 1080p to 2160p is not a small step, and a long 4K download is a genuinely large file that takes as long as your connection needs. The resolution side of that decision is covered in the 4K and quality guide.
Keeping downloads small on purpose
If you only want the sound, taking audio alone is by far the biggest saving available. The video stream is never requested, so for a 1080p source the bytes transferred drop by roughly 10 to 20 times, and in our measurement the MP3 came to 19.0 MB against 161 MB for the 720p video of the same clip.
That difference is mechanical rather than a compression trick. Nothing clever is happening, the picture is simply never fetched. The audio download page is the route for that, and the audio guide covers what quality to expect.
For video, resolution is the dial that matters. Dropping from 720p to 360p in our measurement took the file from 161 MB to 28.8 MB, which is a far bigger saving than any format choice could give you. Converting a file afterwards to shrink it costs quality, because that requires a genuine re-encode, as the converter guide explains.
- Decide what the file is for. Watching on a phone rarely needs more than 720p.
- If you only want the audio, take the audio route and skip the video stream entirely.
- Check the duration, then multiply by a per minute figure to get a rough expectation.
- Pick the lowest resolution that satisfies the purpose, rather than the highest one available.
Estimating before you download
Multiply duration by a per minute anchor and treat the answer as a range. A 40 minute talk at 720p is somewhere around 600 MB using our 15 MB per minute figure, and it could easily land at half that if the content is a static speaker on a plain stage.
Vertical clips are the easy case, because they are usually short. A Short or a Reel measured in seconds will not trouble any storage, whatever resolution you take, which the Shorts guide touches on.
And storage is a different question from where the file goes. On Android the download lands in /storage/emulated/0/Download and appears in the gallery on its own, while on an iPhone it goes to the Files app and needs one extra tap to reach Photos. The mobile guide has both paths.
Frequently asked questions
How big is a ten minute video download?
In our measurement, 161 MB at 720p and 28.8 MB at 360p for a 10 minute 35 second 60fps source. That works out to about 15 MB and 2.7 MB per minute.
How much smaller is audio only?
Much smaller. The same clip came to 19.0 MB as an MP3, about 1.8 MB per minute, and the video stream is never fetched at all.
Why do two videos at the same resolution differ so much in size?
Because bitrate rather than resolution decides size. Frame rate and how much motion is on screen both push bitrate up or down at the same pixel count.
How large is a 4K download?
Considerably larger than 720p. YouTube recommends 35 to 68 Mbit for 2160p30 uploads against 8 to 12 Mbit for 1080p, and delivered streams sit below those upload figures.
Can I shrink a file after downloading it?
Only by re-encoding it, which loses quality. Choosing a lower resolution at download time is the cleaner way to get a smaller file.
Does 60fps really make a difference?
Yes. A 60fps source carries twice as many pictures per second as a 30fps one, and our measured example was 60fps, which puts it at the heavy end.