Add music to a video: the track is the easy half, the mix is the job

Be clear about the search you came from. If the video is finished and just needs a track under it, every phone editor does that job, and none of them charges for it.

Drop the file in, add an audio track, pull the volume down, export. Ten minutes, no new tool, no subscription.

So this page is not for that job, and pretending otherwise would waste your money. It is for the harder case: music under a person talking, in a video that is still being cut.

Under a voice, a track is not decoration. It is a second signal competing for the same ear, and most tracks win that fight by accident.

A spoken voice lives roughly between 500 Hz and 4 kHz, and the consonants that make words legible sit at the top of that band.

A track with energy in the same band masks those consonants. The viewer does not notice that the music is too loud. They notice they stopped following the words, and they scroll.

So a good bed under speech is deliberately dull. Steady, sparse, no vocal, no build that arrives on top of a sentence.

A sung track is the worst offender, because a second set of words in the same range as the first is masking by design. Under speech, instrumental wins almost every time.

You type what the bed should feel like, in plain words. Warm, unhurried, acoustic, nothing dramatic. Or dry, minimal, modern, no build.

Leave the lyrics box empty and the track comes back instrumental, which under a voice is the correct answer.

A bed of up to sixty seconds is 50 credits. A ninety second bed is 90 credits and a two minute bed is 130, priced per started half minute past the base.

Or skip generating and pick from the built-in tracks. In the editor, music is three controls: turn it on, pick the track, set how loud it sits.

At these prices, candidates beat briefs. Generate three moods, sit each under the same cut, and keep the one you stop noticing.

The mix is honest about what it automates. You set one level for the bed and it holds. There is no automatic ducking that dips the music under each phrase.

So set it lower than feels right in a quiet room. On the phone speaker where the video will actually play, that lower level is the correct one.

What is automated is the part nobody should do by hand. Every export is loudness normalised to -14 LUFS integrated with the peaks held at -1 dBTP, inside the render itself.

That means the finished file arrives at the level feeds expect, with no clipped peaks. The voice staying on top is your one decision, the bed level, and the paragraph above is how to set it.

Music lives on Basic: $39.99 a month for 2,500 credits, alongside the avatar module. Lite at $19.99 for 1,250 credits covers the videos themselves.

A thirty second ad with a generated bed is about 160 credits all in. Premium is $79.99 for 5,000 credits, and a $39.99 top-up adds 2,500 when a month runs long.

Now the limit, stated plainly. This writes beds for ads. It does not write the song stuck in your head, and asking for a named artist's sound produces a track you could not safely advertise with anyway.

A bed is furniture under an argument. If the music is meant to be the point of the video, hire a musician for that one and generate the furniture here.