There's been so much buzz around text-to-music generators like Suno and Udio. So, I've decided to put them to a test.
To this day, I still know every word of the 50-second “Education Connection” jingle. Can I recreate that with these generators?
Could one of them produce a catchy, well-produced, break-me-off-a-piece-of-that-Kit-Kat-bar level jingle?
I gave them all the same prompt. Let’s see what they can do.
The Set Up
I first started by identifying a few text-to-music tools to test. After some research, it was clear Suno and Udio are the top options in the market right now.
Other text-to-music generators like Mubert, Soundgen, and Stable Audio all produce background music but no lead vocals. (They still get some honorable mentions later.)
Next, what’s a jingle without lyrics? I went over to ChatGPT for a HubSpot jingle.
Are they amazing lyrics? Probably not, but this is not where I’ll be picky.
From there, I wrote the following prompt to use on all the generators: “Catchy, upbeat and pop jingle with a simple and repetitive melody.”
Generators look for keywords that offer insight into the style and pace of the music you want to produce.
Suno vs. Udio: Which AI music generator produced the better jingle?
I won’t keep you waiting till the end for the answer: It wasn’t even close, Suno outperformed Udio by miles.
Every generator I tested produced 2 clips for every prompt.
Suno’s two outputs, each with a unique sound, both had the pace and style of a jingle. One was very piano forward, a staple in many popular jingles.
The lead vocals were a bit grainy and could use some clarity but they matched the song very well.
Catchy? Check. Good vocals? Check. I give both a solid 7.5/10.
Suno Output 1
Suno Output 2
Udio, however, was…a disaster.
Both the tempo and style were way off and the lead vocals sounded like they were pulled straight from a 13-year old’s Garage Band session.
Udio Output 1
Udio named this song, 'Success in the air.' Unfortunately, this was anything but.
Udio Output 2
To make it worse, this output was v2.
Udio’s first output had inaudible lead vocals mixed with a not-so-great sound. I played around with the levers for prompt and lyric strength to make the lead vocals audible.
Udio v1 Output
Is it saying...words?
In many tests beyond this one, I've noticed Udio struggles to accurately match user prompts with the outputs.
In comparison, Suno can produced a fully customized solid company jingle within minutes and with very little prompting.
Takeaway: While Suno does well with customizations and precise instructions, Udio is ideal for users who aren’t looking for something specific in mind.
Now, let’s move on to some honorable mentions.
Best Text-to-Music AI Generators for Jingles
The following AI generators weren’t able to produce tracks with lead vocals.
However, with the same prompt, they produced high-quality background tracks that could be paired with lead vocals from another software.
Starting off strong with Mubert, which produced the best jingle of all the generators (including Suno). This sound felt modern, catchy and fast paced.
If I was actually producing a jingle, I’d take this background track, upload it to Suno (with a premium account), and layer it with lead vocals.
Soundgen and Stable Audio also produced noteworthy jingles without lead vocals.
Soundgen Output
Okay, drums!
Stable Audio Output
This one sounded like the least modern in the bunch, sounding like a jingle from the early 2000's. Nostalgia is quite a powerful marketing tool, so maybe it would work?
AI Does It Again
With AI music generators like Suno and Mubert, businesses – particularly those with limited budgets – can go beyond what they thought was possible for marketing campaigns.
Most generators outline a few licensing options, including clearance for commercial use at a cost. However, producing the music is absolutely free.
It would be too easy to say, "Here’s something AI sucks at." Alas, it’s not the case here.
![Real or AI-Generated? You Guess [Quiz]](https://www.hubspot.com/hubfs/Real%20or%20AI-Generated.png)
