Wanting to make a listenable song without playing an instrument or knowing any theory was a joke three years ago. Suno turned it into an input box: write a sentence asking for a house track about quitting a job, or hum a melody, and the model fills in the arrangement, the instrumentation and the words. The official pitch is that from a text prompt to a song with vocals and a full arrangement takes under a minute.

From a sentence to a song

The official entrance is restrained. The home page carries a single prompt field, captioned as starting from a short description or moving into the professional editing tool. What separates a beginner from someone experienced are the controls under the prompt: voices, inspiration sources, exclusions, vocal gender, and further down a weirdness setting and a style slider. Those parameters decide whether a track lands near the genre you had in mind or wanders somewhere unexpected.

The lyric layer deserves its own note. Suno does not only generate: it lets you rewrite lyrics, adjust tempo, extend a segment or rework a piece as a cover, all of which the site groups under remix. For people making content, that means no need to throw a version away — you can keep polishing one that already works.

Suno Studio: an audio workstation moved into the browser

The site describes Suno Studio as a new kind of object: a web-based generative audio workstation that puts track-based operations and AI music creation in the same place. The Studio page states it more directly, calling Studio 2.0 control over every detail and the DAW in your browser.

The most practical ability there is stem export: a song can be split into up to 12 time-aligned WAV stems and dragged into a professional audio workstation for further work. For mixers and arrangers this single line decides whether Suno is a toy or a tool that fits a workflow — generation is only the start; whether the material can be handed to professional tools is what matters.

Models and allowances: what the free tier can do

Models are tied to tiers: the free tier uses the free v6-mini model, while paid tiers unlock the two advanced models, v6 and v6-wild. The free tier's limits go beyond the model, and the pricing page lists them plainly: 50 credits a day, at most 10 songs, standard features only, audio uploads of no more than 8 minutes, a shared queue, no monthly downloads, no commercial rights, no credit top-ups and no stem separation.

Paid tiers differ in more than volume. From Pro there is a priority queue, up to ten songs generated in parallel, the ability to record or upload your own voice, and training of custom models. Premier adds full access to Suno Studio — MIDI, effects, automation and custom plugins — lifts the upload ceiling from 8 minutes to 30, and allows adding vocals or accompaniment to a song that already exists.

The quotas behind that: Pro carries 2500 credits and 20 downloads a month, Premier 10000 credits and 60 downloads. On price, the annual switch on the site shows Pro at 8 dollars a month billed yearly and Premier at 24 dollars, with the page noting savings of 24 and 72 dollars. Yearly is roughly a fifth cheaper, but for a short trial the monthly option is more flexible.

Commercial rights: check this first

This is where people trip: commercial rights start only at Pro. Songs generated on the free tier carry no commercial rights according to the site — you can listen to them and send them to friends, but you cannot put them under a video, use them as advertising music or release them.

For Pro and above the site says songs created by a paying subscriber belong to the subscriber and can be used freely, including as video background music or on a released album. For independent creators, musicians and brand content, that one line weighs more than the parameter table.

Not only generation, also a community

The site puts sharing on the same level as generating: a finished song can be published straight away, from close friends to a wider audience. The browse page carries the other half — you can listen to what others made to find a direction while posting your own covers, reworks and experiments. For a newcomer that beats reading a tutorial: seeing what someone else produced from the same prompt gets you to the edges of the parameters quickly.

Who it suits

If what you need is usable original music — background for short video, a podcast intro, a looping track inside a game, a theme for an internal event — Suno's output rate is hard to match by hand. Such needs used to mean buying a stock library or hiring a writer; now you can generate in bulk and then polish the few that fit.

If the goal is a piece that carries a personal style and survives repeated listening, it is a different story. Generated songs tend to be very standard in structure and emotion, and what they lack is exactly the accident and the flaw. The more practical use is as an arrangement sketchpad: build the skeleton here, take the stems into a workstation, and turn it into a real piece with your own playing and mixing.

One thing to confirm beforehand: for work involving voices and timbre, especially uploading your own voice or imitating a particular singer, settle the platform rules and the boundaries of authorisation first. Copyright practice in music is still evolving, and the earlier the chain of permissions is clear, the fewer problems later.

What AI music tools really change is the floor of creation, not the ceiling. They make having a first version almost free, and that is exactly where most people get stuck — first have one, then talk about making it good.