Change Video Speed

Speed up or slow down MP4, MOV, M4V and WEBM in your browser. Five presets from 0.5x to 2x, audio pitch preserved, nothing ever uploaded.

🌐 Español

Drop your file here (.mp4, .mov, .m4v, .webm)

🔒 Private by design: your files are processed locally in your browser and never uploaded to any server.

Written and maintained by Last updated

Presentation timestamps, not invented frames

A video file is a pile of compressed images plus a schedule saying when each one should appear. Changing speed here rewrites the schedule. The filter divides every presentation timestamp by the speed factor, so at 2x each frame is due half as early as before and the clip finishes in half the time, while at 0.5x the timestamps stretch and everything takes twice as long.

The frames themselves are untouched in number. Nothing is generated, interpolated or blended, which is the single most useful thing to understand before choosing 0.5x. A 60 fps source slowed to half speed still delivers 30 distinct images per second of the new running time, so it stays smooth. A 24 fps source treated the same way delivers 12, and on fast motion you will see each one. Slow motion that looks cinematic comes from a camera that recorded extra frames; slow motion made afterwards is a redistribution of what already exists.

Sending a clip through

  1. Drop a video on the box or click Choose a file. The accepted extensions are .mp4, .mov, .m4v and .webm; anything else is rejected on the spot with a message naming the file and listing what is allowed.
  2. Choose a Speed. There is no 1x entry, because a pass that changed nothing would still cost you a full re-encode.
  3. Click Change Video Speed. A progress bar appears; Cancel abandons the run and empties the dropzone, so a cancelled job leaves nothing behind to resume.
  4. Download the MP4. It keeps your original name with the factor appended, as in demo-1.5x.mp4.

The first run in a session pulls the FFmpeg core down before it can start, so give it a moment and a working connection. After that it stays loaded for the rest of the tab’s life.

The preset list ends at 2x for a filter-shaped reason

Audio cannot simply be replayed faster without the pitch rising with it, so the audio track goes through a tempo filter that changes duration and leaves pitch alone. That filter accepts a factor between 0.5 and 2.0 and rejects anything beyond it. The five presets here, 0.5, 0.75, 1.25, 1.5 and 2, sit exactly inside that window, which means the audio never needs to be passed through more than one tempo stage. Fewer stages means fewer opportunities for the splicing artefacts that time-stretching can introduce.

If you need to go further than double speed and only care about the sound, the audio-only Change Audio Speed tool reaches 3x by chaining stages and returns an MP3 rather than a video.

Silent clips take a different path entirely

Screen recordings, muted exports and phone clips captured with the microphone off frequently carry no audio track at all. Before encoding, the tool inspects the file’s stream list and picks its filter setup from what it finds. With audio present, video and sound are re-timed in parallel and remapped into the output. With no audio present, only the picture is filtered and audio is switched off explicitly, which also cleans up containers carrying an empty or data-only stream. Nothing is asked of you either way, and there is no checkbox to get wrong.

Every output is H.264 in an MP4

Re-timing the frames rules out a lossless path. A trim or a rotation can often be done by rewriting the container around the existing compressed data, but rescheduling every frame means decoding and encoding the picture again, so a generation of quality goes regardless of which preset you pick. The guide to what re-encoding costs is worth reading if you plan to stack several edits on the same clip.

The practical advice that follows is to make the speed change the last step, or at least a late one. Do the lossless work first: the Trim Video tool cuts to a start and end time with a stream copy and never re-compresses a frame, so trimming first means the single encode runs over less material. And if what you actually want is a different frame rate at the same running time rather than a different running time, Convert Video Frame Rate is the right tool, since it duplicates or drops frames to hit a target rate and leaves the clip’s duration alone.

Frequently asked questions

Is 0.5x the same kind of slow motion a phone shoots at 240 fps?

No, and the difference is visible on fast movement. A phone shooting high frame rate captures four or eight times as many real moments and then plays them at a normal rate, so every instant is genuinely recorded. This tool has only the frames your camera already captured and spreads them over a longer period, so a 30 fps clip at half speed shows 15 distinct images per second of playback. Slow pans and steady movement hold up fine. A fast hand gesture or a ball in flight will look stepped, and no setting fixes that because the missing images were never recorded.

My source had subtitles and chapters, and the MP4 does not. Why?

The command builds an output from exactly two streams, the re-timed video and the re-timed audio, and maps only those into the file. Anything else the container was carrying, subtitle tracks, chapter markers, data streams, attached fonts, has no place in that mapping and does not survive. It is a real limitation rather than an oversight. If a clip has burned-in captions they are part of the picture and come through as normal, since they are pixels rather than a separate track.

Why does it look like my file is being read twice?

Because it is. Before any encoding happens the tool runs a quick inspection pass over the input purely to read its stream list, so it can find out whether an audio track exists at all. The answer decides which of two different filter setups is used. Referencing an audio stream that is not there would abort the whole encode, so the check is cheap insurance. The inspection pass is fast compared with the encode that follows it.

What quality does the re-encode use, and can I change it?

The video is encoded with H.264 at a constant-rate-factor of 20 using the fast preset, and the audio with AAC at 128 kbps. None of that is exposed as an option on this page, which is a deliberate simplification. On the CRF scale a lower number means higher quality, and 20 is a step better than x264's own default of 23, close enough to the source for most material. The fast preset trades a little compression efficiency for a shorter wait, which matters more than usual when the encoder is running single-threaded inside a browser tab.

Can I put a WEBM in and get a WEBM back?

No, the output is always MP4. Since a speed change forces a full re-encode anyway, there is no original stream worth preserving in its original container, and MP4 with H.264 and AAC is the combination that plays without argument on phones, editors, social platforms and older desktop players. If you specifically need WEBM back, convert the finished MP4 afterwards with a dedicated container converter.

The finished screen shows an arrow between two file sizes with a percentage. What is it comparing?

The total size of what you dropped in against the total size of what came out, with the percentage being the difference between them. It is not a quality measurement. A clip sped up to 2x usually shrinks a lot simply because half the duration means half the frames to store, while a clip slowed to 0.5x often grows for the same reason in reverse. Reading that number as evidence about how good the encode looks would be a mistake.

Related tools