Compress Video

Target size is for when a limit is printed somewhere: an email provider, a chat app, a job portal or a course platform. You type the number, choose MB or KB, and the compressor works backwards from the length of the clip to the bitrate that fits. The line above lists a few published limits from email and file-sharing services; check the current figure on the service you are sending to, because these numbers change.

Quality level is for when there is no hard limit and you just want a lighter file. High keeps the picture closest to the original, Balanced is the everyday choice, and Small and Smallest give up more detail for lighter files. The final size is not fixed in this mode; it follows from the content.

Resolution scales the picture down to a height between 1080p and 240p, with the width following the original shape. A phone screen shows little difference between 4K and 1080p, so lowering the resolution is often the cleanest way to reduce a phone video by a large amount. This mode is also the right one when a platform states a maximum resolution.

A compressed video is a stream of data spread over time. Its size is roughly the bitrate multiplied by the duration, plus a small overhead for the container. When you ask for a fixed size, the duration is the one thing the compressor cannot change, so a longer clip gets fewer bits for every second. Twice the length at the same target means half the bitrate, and the picture shows it.

Motion matters too. A talking head against a plain wall, a slide presentation or a screen recording changes little from frame to frame and compresses very well. Sports, gaming, handheld footage, rain, confetti and crowds change everywhere at once and need far more data to look clean. Two clips of the same length and size can look very different for that reason.

Sound takes a share of the budget as well. The default Audio bitrate of 128 kbps is fine for music; 96 kbps or less is enough for speech. For very small targets, every kilobit spent on sound is taken from the picture, so lowering the audio bitrate or ticking Remove audio for a silent clip gives the video more room.

A practical way to judge a result is to watch the busiest few seconds of the compressed copy at full screen, on the kind of screen your viewer will use. If faces, text and edges still look clean there, the rest of the clip will look fine as well. If they do not, raise the target, lower the resolution one step, or cut the clip shorter, and compress the original again.

Most of the controls under Advanced settings trade one thing for another. Maximum resolution caps the height in target-size and quality modes, which is useful when a 4K phone recording has to become a modest file. Maximum frame rate can bring 60 fps down to 30 or 24; the change is hard to see in calm footage and easier to see in fast pans. Video codec decides how efficiently the picture is stored, and Two-pass encoding spends extra time analyzing the clip so the size lands close to your target.

Remove audio and Audio bitrate sit in their own group. For a screen recording with narration, 64 or 96 kbps keeps the voice clear and frees space for the picture. Strip metadata is on by default and removes the location and device details a phone writes into every recording, which is worth keeping on for anything you send to strangers. None of these settings touch the original file on your device; you always get a new copy, so it is safe to try a setting, look at the result and run it again with a different one.

The codec is the method used to store the picture, and newer ones are more efficient. H.264 is the default because nearly every phone, computer and TV in use today plays it without extra software. H.265 (also called HEVC) and AV1 store the same picture in less data, which helps most at small targets, but older Windows laptops, budget Android phones, smart TVs and some browsers cannot play them. VP9 is mainly used with WebM on the web. MDN's video codec guide compares them in more depth.

Output format offers MP4, WebM, MOV or the same format as the source. MP4 is the one to pick when the file goes to another person or into an app you do not control. WebM suits a page on your own website, and MOV makes sense when the file goes straight into an Apple editing app. Keeping the source format is handy for batches of MKV recordings you want to archive at a smaller size without changing how they are organized.

Some limits are better solved another way. Large video platforms accept big uploads and encode everything again on their own servers, so compressing beforehand only removes detail they could have used; upload the original if the platform allows its size. Cloud storage links are often the easier route for long recordings that need to stay sharp, since services such as Google Drive and Dropbox accept files far larger than any email attachment.

If the problem is that a file will not open rather than that it is too big, converting is the better fix. The Video Converter page changes the container and codec without aiming for a size. And if only the sound matters, as with a recorded lecture or a voice message, Video to Audio keeps just the audio track, which is a small fraction of the original file.

For a short loop that will be posted in a chat or forum, consider the format as well. A few seconds saved as MP4 is usually lighter than the same few seconds as a GIF, because GIF stores every frame as an image. Many apps that accept GIFs also accept small MP4 clips and play them the same way.