Which section tags does Suno actually follow
[Intro], [Verse], [Pre-Chorus], [Chorus], [Bridge] and [Outro] are a common convention and most tools follow them often but not always. Put them on their own line directly above the lyrics for that section. Some tools also accept [Instrumental] to mark a section with no vocals.
- [Intro]: usually short, sets the mood before words start
- [Verse]: the storytelling section, changes each time it repeats
- [Pre-Chorus]: a short lift that builds tension into the chorus
- [Chorus]: the hook, should repeat with near-identical wording
- [Bridge]: a contrast section, often once near the end
- [Outro]: closes the song, can repeat the chorus or fade
How many lines per section
Four lines per verse and four per chorus is a reliable default for a 3 to 4 minute pop song. Shorter sections (2 to 3 lines) work better for a 30 to 90 second clip, since the model has less time to develop each part.
| Section | Typical lines |
|---|---|
| Intro | 0 to 2 (often instrumental) |
| Verse | 4 |
| Pre-Chorus | 2 |
| Chorus | 4 |
| Bridge | 2 to 4 |
| Outro | 0 to 2 |
Too many syllables packed into one line is a common cause of rushed or mumbled vocals, so keep lines short enough to be sung comfortably.
How do style tags and lyrics tags work together
The style or prompt box carries the sound (genre, tempo, instruments, mood, voice), and the lyrics box carries the words and structure tags. Keep the style prompt to 4 to 7 descriptors: an anchor genre, drums, bass or low end, lead instrument or voice, mix character, and one mood word.
For example, pair the tag structure above with a style prompt like "pop, upbeat, warm analogue synth, punchy kick and clap, bright female lead vocal, layered backing vocals in the chorus, 116 BPM" rather than repeating structural words like "verse" or "chorus" inside the style box, since that box is read as sound, not layout.
Making the chorus repeat as a real hook
Write the chorus lyrics identically every time they appear in the lyrics box, since small wording changes between repeats can make the model treat each one as a new section rather than a returning hook. Keep the chorus melcially simple in wording too: short phrases, a repeated word or two, and a clear rhyme land better than a chorus packed with new information.
If the hook still comes out weak, try moving one line from the chorus into the pre-chorus so the chorus itself carries fewer, punchier words.
Lyrics skeleton
Original, for structure
[Verse] Streetlights blink on, the city exhales I'm folding your name into paper trails One more corner, one more stop Before the feeling starts to drop [Chorus] Hold the light, hold the line Say it slow, make it mine Hold the light, hold the line We've got time, we've got time
Common mistakes
- Putting section tags in the style box instead of the lyrics box, which gets ignored: keep tags with the lyrics.
- Rewriting the chorus slightly each time it repeats, which weakens the hook: reuse the exact same chorus lines.
- Writing 8 or more lines per verse for a short clip, which forces rushed vocals: cut to 4 lines for anything under 90 seconds.
- Mixing structure words like 'chorus' into the style prompt, which the model reads as sound description not layout: keep the style box for genre, instruments, voice and tempo only.
- Skipping the pre-chorus and jumping straight to the chorus, which can make the hook land flat: add a short 2-line pre-chorus to build lift.