AI jingle for your brand: cheap to make, expensive to change

The lyrics box takes the words. Keep them under about fifteen, because a jingle has to survive being half heard.

The description box takes the sound in your own words. Upbeat pop, bright and catchy, female lead vocals is a complete brief.

Pick 15 seconds. Thirty is available and it is usually too long for something meant to be recognised, not listened to.

The price is 20 credits, flat, at every length. So fifteen attempts costs 300 credits, which is exactly the free tier with no card.

Generate five, sleep on it, then generate five more. A jingle you like on the first listen is often the one you tire of by March.

There is no editing after the fact. If it is close but wrong, adjust the description and spend another 20 credits.

A jingle works through frequency. It becomes recognisable somewhere around the twentieth exposure, and it means nothing before that.

Which means a brand that changes the music every campaign never gets any value from having made one at all.

The decision is not what should it sound like. It is are we willing to run the same fifteen seconds for a year.

If the answer is no, generate ordinary background music per ad instead and save yourself the exercise.

If the answer is yes, the 20 credits is the cheapest part of the commitment by an enormous margin.

Everything expensive about a jingle is the discipline of not changing it when the team gets bored, and teams get bored around month four.

Write the commitment down before you generate anything. A line in a document saying this runs until next March is worth more than the track it protects.

Somebody is scrolling. They hear roughly three words before deciding whether to keep watching.

So the brand name goes in those three words, or the jingle is doing nothing a bed would not do more cheaply.

Keep it under fifteen words total. A jingle is not a verse and it is not a chorus, it is a phrase with a tune attached.

Avoid anything that needs to be understood. Nobody is parsing a clause while deciding whether to scroll.

Say the name, say what you do, stop. Fifteen words is generous for that.

Then say it out loud twenty times. If you are sick of it, that is the correct feeling and it is not a reason to change it.

A sung vocal and a spoken voice compete for the same frequencies. They compete for the same attention too. The viewer follows neither.

The pattern that works is the jingle in the opening two seconds, then an instrumental bed under the talking.

That is two tracks at 20 credits each. Describe the bed as sparse, no vocals, nothing in the midrange, and it will stay out of the way.

In the editor the music controls are turn it on, pick the track, and set how loud it sits. No ducking and no stem separation.

So the separation has to be structural rather than technical: the jingle owns the top, the bed owns the rest.

Check it on a phone at half volume. That is where a sung line and a voice collide if they are going to.

Look at what music your last ten ads used. If it was ten different beds, you are not a brand that keeps a jingle yet.

That is not a criticism, and it costs you nothing here. A fresh bed per ad is the same 20 credits and no commitment at all.

Three patterns say wait. Fewer than twenty ads a year means nobody hears it enough times for recognition to build.

Positioning that is still moving means a jingle locks in a claim you may want to change.

A considered purchase decided over three weeks of research gives a fifteen second tune almost nothing to do.

If none of those apply, make it, and hold it. The 20 credits was never the expensive part.

Three plain limits sit alongside. No editing after generation, no stems, and 120 seconds is the ceiling on any track.

The track is generated for you rather than licensed. Confirm the usage terms that apply to your own account before making one a permanent brand asset.