For decades, the music industry has been defined by high barriers to entry. To release a professional record, you generally needed years of instrument mastery, expensive studio time, a sound engineer, and a record label to back you up.

But as we close out 2025, the GenAI revolution has completely dismantled those gates. Over the holiday break, I decided to put the latest technology to the test to see if I could go from concept to published artist without an agent or a recording studio.

The result is my debut EP, Vibe Code to This, and the process behind it is a testament to how accessible creativity has become. This isn’t just about fun experiments anymore, we are now at a point where AI tools can produce commercial-grade creative assets that rival human production.

Suno v5 offers stunning capabilities

Core to this project was the music service Suno, specifically their latest v5 Large Language Model. If you have played with text-to-audio tools in the past, you might remember them being low-fidelity or lacking structure, but v5 is a massive leap forward.

A year ago, I played with Suno and found it technically interesting, in the way it could create songs from text, but the output was fun and entertaining, not of a quality that was something I’d put on in the car to listen to while driving. Thankfully, they kept investing, and things have changed a lot.

Suno v5 understands musical theory, vocal cadence, and genre-blending in a way that feels shockingly intuitive. The platform operates on a credit system, but for the quality of output you get, the cost is negligible compared to hiring session musicians.

The process starts with a text prompt, but it requires more than just typing “make a cool song.” To get the specific tech-themed aesthetic I wanted, I had to be precise with my instructions and refine it time and time again, but within a short period of time, I had something I was pretty happy with.

When you create in Suno, you need to provide 3 things. You need to define the style of the music, the lyrics in the song and the title for the song.

This year, I’ve gone back and forth between Grok 4.x and Gemini 3.x, and for this task, I turned to Gemini thanks to their Gem feature. This allows me to store an instruction set and prompt against it, speeding up tasks that are likely to be repetitive (like making multiple songs).

Having defined what Suno needed and the overall theme of what I was chasing, I built a MusicGen Gem in Gemini, then got to work. I then took the output from Gemini and copied that into Suno and hit Create. A few seconds later, I had 2 options to choose from.

I would listen to both, analyse what I liked, what I didn’t, then update the prompts and run through the cycle over and over. While this took some time, it was incredibly fast compared to traditional music creation. This felt like a superpower.

The prompts specified styles like “driving synthwave,” “staccato basslines at 128 BPM,” and “emotional female vocals.” The model’s ability to adhere to these constraints was impressive, and after some time in the game, I grew more confident in crafting the prompt to get the style of song I wanted.

The generated lyrics also took an equal amount of iterative cycles to achieve. We don’t find Technology references in songs very often, and the model wants to rhyme everything, so it was common to get suggestions to change words or acronyms for the betterment of the song, but then could make something illogical or technically incorrect. This was a fun challenge to work through, and an upcoming song on the second Album, called ‘The Cold Aisle’, is another great example of this.

From generation to distribution

Creating the audio files is only the first half of the battle. To actually be an “artist” in the modern sense, your music needs to live where the listeners are, Spotify, Apple Music, YouTube Music, and Tidal.

Historically, you couldn’t just upload MP3s to Spotify directly, you needed a distributor. This is where the industry used to force you to engage management agencies and sign contracts with labels, but the tech sector has solved this too.

I used a platform called Amuse.io to handle the distribution. It is a streamlined, app-based record label that allows independent artists to push their tracks to all major streaming services without signing away their life rights.

Amuse is one of a number of services listed on the Spotify for Artists page, giving you confidence it’s legit, and of course, you don’t have to use this one, select any on this list, each offering different pricing.

The interface is incredibly clean and tech-forward. You simply upload your lossless audio files (WAV or FLAC is essential for quality), upload your cover art, and fill in the metadata.

To complement the music, you’ll need some graphics, Cover Art for your track, album, artist profile pic and header backgrounds for the streaming services. For this, again, I turned to GenAI, leveraging Google Gemini Nano Banana to take the lyrics from a song and create a relevant image for the single. This again took iteration, and over time, I got better and faster at prompting, even creating another Gem to handle the image generation task I would to use for each song.

On paid plans, Suno offers the ability to download in .WAV format, and higher tiers offer even higher quality outputs. Just download the Audio, upload to a Release in Amuse, fill in details about the music rights (if you did all the work, this bit is quite easy).

I opted for the ‘Pro’ tier to ensure the release happened on my timeline. The cost was roughly A$35.99 per year, which is a tiny investment for global distribution capabilities.

Retaining rights and royalties

One of the most critical aspects of using a service like Amuse is the ownership structure. In traditional music deals, the label often owns the master recordings.

With this modern tech stack, I retain 100% of the rights and 100% of the royalties. When you stream a track from the EP, the revenue flows directly back to the creator, filtered through the Amuse dashboard, where you can track streams and analytics in real-time.

This transparency is something the music industry has desperately needed for a long time.

Having now had the music live for a number of days, the service is now starting to show aggregated listening analytics, while also showing a breakdown of listeners from different services, allowing you to focus promotion or investments based on data.

The democratisation of creativity

What this project proved to me is that the definition of a “creator” is expanding rapidly. I am not a singer, nor can I play the drums, yet I was able to produce an album that sits comfortably alongside human-produced electronic music.

Critics often worry that AI will replace artists, but I see it differently. These tools unlock creativity for people who have the vision but lack the technical motor skills to execute it on a physical instrument.

It allows us to focus on what you want the music to feel like and if you care, the specifics of the lyrics, to create something truly unique. The heavy lifting of compression, mixing, and mastering is handled by the algorithm, freeing the human to focus on the direction and the story.

The timeline from “I have an idea for a song” to “this song is on Spotify” has compressed from months to mere days and I have done exactly that, had an idea, run it through this worklow I’ve created and had a song I’m proud of, that I can share with friends, family of the entire world in an incredibly short period of time.

Final thoughts

If you have ever had an idea for a melody or a concept for an album, there has literally never been a better time to execute it. The tools are affordable, the quality is professional, and the gatekeepers have been removed.

My debut EP is live now, and while I hope you enjoy listening to it, I hope even more that it inspires you to open up these tools and see what you can build yourself.

“Everything’s binary, black and white. But I’m seeing colours in the neon light.”

For more information, head to https://techau.com.au/techau-launches-debut-ep-vibe-code-to-this-for-the-late-night-creator/