How to make your first AI song step by step

To make your first AI song, open the AI Music Generator, decide whether you need vocals, choose Simple Mode or Advanced Mode, describe one clear musical goal, and select Generate Music. The first result is a draft to evaluate, not a final song you must accept.

Before touching the controls, finish this sentence:

I want a song for ______ that feels ______, led by ______.

For example:

I want a song for a late-night travel video that feels wistful but hopeful, led by fingerpicked acoustic guitar.

That sentence gives you a listening target. Without it, every generated result can sound interesting while still being wrong for the job.

Choose Simple Mode or Advanced Mode

Start with Simple Mode when you have an idea but no fixed lyrics. It gives you one required Description field, so you can explain the whole song in natural language.

Choose Advanced Mode when you need separate control over:

• Title;
• Style;
• Lyrics;
• optional Advanced Options when they are available to your account.

In either mode, select With Vocals for a sung result or Instrumental for music without a vocal part. In Advanced Mode, Title is required. Lyrics are also required when With Vocals is active; switching to Instrumental hides the lyric field.

If you are new, Simple Mode is usually the fastest way to learn what your brief actually means in sound. Move to Advanced Mode when you know which details need separate control.

Make the first version in Simple Mode

Use this sequence:

1. Select Simple Mode.
2. Choose With Vocals or Instrumental.
3. Select an available AI Model Version. Availability can vary by account, so use the choices shown on your screen.
4. Write the purpose, genre family, emotion, movement, and one focal sound in Description.
5. Select Generate Music.

A useful description is compact but testable:

Intimate acoustic indie folk for a late-night travel video. Wistful but hopeful, unhurried at first, then gently building around fingerpicked guitar and a low, close vocal.

This brief answers five practical questions: where the song will be used, its musical family, its emotional direction, how its energy moves, and what should lead the arrangement. For more examples, see how to write a simple AI music prompt.

Avoid writing a pile of unrelated adjectives. “Epic, dreamy, cinematic, catchy, dark, happy” does not tell you which quality should win when the words conflict.

Use Advanced Mode when lyrics or structure matter

In Advanced Mode, enter a short Title, then use Style for the musical and production direction. If With Vocals is selected, put only the words to be sung and useful section labels in Lyrics.

A practical setup looks like this:

Title: Midnight Windows
Style: intimate acoustic indie folk, restrained drums, close low vocal,
       narrow verse, wider final chorus
Lyrics:
[Verse]
Streetlights slide across the glass...

[Chorus]
I keep the road inside my name...

The Open Professional Lyrics Editor control opens the larger writing workspace. Use Quick Tags to organize sections, revise the draft, and save it back before generating. If the words need significant work, follow the professional lyrics workflow first.

Open Advanced Options only when you can name a specific problem. If an option or model is unavailable, do not rewrite the tutorial around it; account access and model availability can change.

Listen to the result with four questions

When generation finishes, compare the returned results against the original goal:

1. Does the intended emotion arrive early?
2. Is the vocal or lead instrument appropriate?
3. Does the chorus or central section create enough contrast?
4. Is there one performance, texture, or section worth protecting?

Record one short note per result:

Version A: right guitar texture; chorus arrives too early.
Version B: better pacing; vocal is more theatrical than the brief.

Change only the most important failure in the next attempt. If you switch mode, rewrite the lyrics, change the model, and add several controls at once, you will not know which decision helped.

Fix common first-song mistakes

• The result feels generic: replace vague praise words with a use case, emotional direction, and focal sound.
• The song has vocals when you wanted background music: use Instrumental instead of asking for “no vocals” inside a vocal brief.
• The lyrics are ignored or incomplete: use Advanced Mode, keep With Vocals active, and put the actual sung words in Lyrics.
• Every attempt sounds unrelated: keep the base brief stable and change one variable at a time.
• You are chasing a perfect take: save the strongest section first, then decide whether the next step is another generation or an edit.

Know when to move from generation to editing

Stop regenerating when the track has a convincing identity, at least one strong section, and a performance or groove you would be unhappy to lose. At that point, editing protects more value than another complete reroll.

The current studio can continue a promising track with Extend and can replace an eligible timed lyric passage with Edit Music. Read how to extend and complete a song before changing a track that already contains something worth keeping.