Free tool · MP4 · Runs in your browser
Add audio to a video
Short answer
Choose a video and an audio file, decide whether the new sound replaces the original or mixes with it, set the volume, and download an MP4. The picture is copied without re-encoding, so it is untouched (the video stream had the same checksum in all nine of our tests) and the job took 0.2 to 0.4 seconds on a 5.1 second 1080p clip. The result is always as long as the video: longer audio is cut at the end, shorter audio repeats if you tick the loop box, or is followed by silence. Nothing is uploaded.
Maximum 100 MB. Both files stay on your device. The picture is copied without re-encoding; the result is as long as the video.
Key takeaways
- The picture is not re-encoded: its checksum was identical before and after in 9 of 9 runs, so there is no quality loss and the file size changed by only -0.10% to 7.60% (the soundtrack's size).
- The new audio is encoded as AAC at about 191 kbps, the format MP4 players expect, whatever the audio file was (an MP3 in our tests).
- The result is as long as the video: a 30 second track on a 5.1 second clip gave 5.1 seconds, not 30.
- A track shorter than the video: with the loop box off, the last 1.5 seconds of our 3 second track on the 5.1 second clip were silence (silence); with it on, they were -25.9 dB, the same level as the start.
- Volume is in percent of the audio's own level: 50% played -6.1 dB and 200% +6.0 dB compared with 100%.
Measured: what each setting does
Measured with the tool's own ffmpeg arguments on the same single-thread WebAssembly engine your browser downloads, on a 5.1 second 1920 × 1080 excerpt of Big Buck Bunny (8.2 MB, with a speech and street noise soundtrack at -22.1 dB mean level) and excerpts of a public domain recording of a Mozart piano sonata (30 seconds and 3 seconds, mean level -26.0 dB). The levels are the mean volume reported by ffmpeg's volumedetect filter over the first 2.5 seconds and over the last 1.5 seconds of the result. Times are from a headless run; your device will differ.
| Case | Time | File size change | Result length | Level, start | Level, end | Picture identical |
|---|---|---|---|---|---|---|
| Replace the soundtrack with longer music | 0.42 s | +0.5% | 5.08 s | -26.1 dB | -22.1 dB | Yes |
| Replace with 3 second music, no loop | 0.26 s | -0.1% | 5.11 s | -26.1 dB | silence | Yes |
| Replace with 3 second music, loop | 0.31 s | +0.5% | 5.07 s | -26.1 dB | -25.9 dB | Yes |
| Mix longer music at 30% with the original | 0.23 s | +0.3% | 5.07 s | -22.8 dB | -21.1 dB | Yes |
| Mix longer music at 100% with the original | 0.19 s | +0.4% | 5.07 s | -20.6 dB | -17.9 dB | Yes |
| Silent video, add music | 0.32 s | +7.6% | 5.08 s | -26.1 dB | -22.1 dB | Yes |
| Silent video, mix selected (no soundtrack to mix with) | 0.30 s | +7.6% | 5.08 s | -26.1 dB | -22.1 dB | Yes |
| Replace, music at 50% | 0.36 s | +0.5% | 5.08 s | -32.2 dB | -28.1 dB | Yes |
| Replace, music at 200% | 0.32 s | +0.5% | 5.08 s | -20.1 dB | -16.1 dB | Yes |
Every case, with the tool's default of AAC at 192 kbps and a copied picture.
The result is 0.02 to 0.04 seconds longer than the picture in the replace cases because AAC audio is written in frames of 1,024 samples. The mixed results are exactly as long as the picture.
Replace or mix?
Replace drops the original soundtrack and uses only the new audio. Mix keeps the original and adds the new audio on top, with a separate volume for each; the filter does not rescale the levels, so loud sources add up. On our clip the original soundtrack was at -22.1 dB and the music at -26.0 dB.
| Setting | Level at the start of the result | Against the original soundtrack |
|---|---|---|
| Original soundtrack alone | -22.1 dB | 0.0 dB |
| Replace, music at 100% | -26.1 dB | -4.0 dB |
| Mix, music at 30% over the original | -22.8 dB | -0.7 dB |
| Mix, music at 100% over the original | -20.6 dB | +1.5 dB |
For background music under speech, mix with the music at 20% to 40% and keep the original at 100%, then listen: the right level depends on how loud the music is recorded. Adding two full-level sources raises the total level, so lower one of them if the result sounds harsh.
When the audio and the video are different lengths
| Audio | Loop box | Result length | Last 1.5 seconds |
|---|---|---|---|
| Longer (30 s) | Either | 5.08 s | -22.1 dB |
| Shorter (3 s) | Off | 5.11 s | silence |
| Shorter (3 s) | On | 5.07 s | -25.9 dB |
FFmpeg's documentation gives the pattern used here for a video that should always run to the end: pad the audio with silence and use -shortest, so the video stream is the shortest. With looping, the audio input is repeated without end and the same option stops the result at the video's end.
A video with no sound
A silent video has no soundtrack to keep, so mix and replace give the same result. Our silent test clip took 0.32 seconds and its file grew by 7.6%, the size of the new soundtrack. If you choose mix for a silent video, the tool notices that there is nothing to mix with and adds the audio on its own.
Tips
- Use music you have the right to use. Platforms can mute or remove a video with copyrighted music.
- Fade or trim the audio first with the audio trim tool if you want it to end at a particular second.
- If the video's own sound is noisy, clean it with the noise remover before mixing.
- Keep the original files: the tool makes a new MP4 and never changes your files.
Free tools:Trim audioNormalize audio loudnessRemove background noise from videoMute a videoTrim a video
What happens inside
The tool runs ffmpeg compiled to WebAssembly in your browser tab. The video stream is copied as it is, the new audio is scaled to the volume you chose, padded with silence (or repeated, if you chose loop), mixed with the original soundtrack if you chose mix, and encoded as AAC at 192 kbps. The output stops when the video ends. Because the picture is copied rather than re-encoded, the job takes a fraction of a second on a short clip; the work grows with the file size, not with the length of the audio.
Questions people ask
How do I add music to a video?
Choose the video and the music file, pick Replace or Mix, set the volume and press the button. Download the MP4.
Does it re-encode my video?
No. The picture is copied as it is, so there is no quality loss. Only the soundtrack is encoded, as AAC at 192 kbps.
What if the music is longer than the video?
It is cut where the video ends. The result is as long as the video.
What if the music is shorter than the video?
Tick the loop box to repeat it until the video ends; without it, the rest of the video is silent.
Can I keep the original sound and add music?
Yes. Choose Mix and set the volume of each. For music under speech, 20% to 40% for the music is a good place to start.
Which audio formats can I use?
MP3, M4A, WAV, AAC, OGG and FLAC all work in the browser engine; the result is always AAC in an MP4.
Is my video uploaded?
No. Both files stay on your device.
What is the size limit?
Up to 100 MB per file.
Related
Sources and how we checked
- FFmpeg Filters Documentation: amix, apad (read 2026-10-09)
- FFmpeg Documentation: ffmpeg (read 2026-10-09)
- Wikimedia Commons: Mozart, Piano Sonata No. 10, K. 330, 3rd movement, Vadim Chaimovich, Musopen (public domain), used as the test music (read 2026-10-09)
- Blender Foundation: Big Buck Bunny (CC BY 3.0), the clip used for the measurements (read 2026-10-09)
Reviewed October 9, 2026. The measurements ran the tool’s own ffmpeg arguments on the same single-thread WebAssembly engine your browser downloads. The script, the results and the saved documentation are in data/seo/hub/evidence/2026-10-09/tools.