Write Song Lyrics That Work in Suno

What does this setup do?

Draft structured, Suno-ready lyrics with section tags, vocal cues, and a matching style prompt.

Who is it for?

Hobbyists making AI music who want songs with real structure instead of mush.

How do you set it up?

Tools

• ChatGPT or Claude (free tier is fine)
• A Suno account (Custom Mode for lyrics input)
• An idea: theme, genre, and mood

Steps

1. Decide the song's arc first: sparse intro, rising pre-chorus, big chorus. Structure is what separates a song from a loop.
2. Ask AI for lyrics with Suno section tags: one tag per line in square brackets directly above that section's lyrics, never inline. Core tags: [Verse], [Pre-Chorus], [Chorus], [Bridge], [Outro].
3. Keep verses 4 to 8 lines and choruses 2 to 4 lines. Do not repeat the chorus more than 3 times. Put your strongest line first in each section.
4. Add vocal delivery cues in parentheses where it matters: (whispered), (belted), (spoken word), (harmonized), (ad-lib), (falsetto).
5. Write the style prompt separately: 8 to 15 comma-separated tags covering genre, mood, vocal direction, instrumentation, and production or tempo.
6. Paste tags-free production notes into Suno's Style field and the tagged lyrics into the Lyrics field. Prose directions left in the Lyrics field can get sung by the model.

Write Suno-ready song lyrics. Plain and practical.

Theme: [what the song is about]
Genre and mood: [e.g. indie folk, bittersweet]
Perspective: [e.g. first person, looking back]

Write:
1. Full lyrics with Suno section tags, one tag per line in [square brackets] directly above each section: [Verse], [Pre-Chorus], [Chorus], [Bridge], [Outro].
2. Verses 4-8 lines, chorus 2-4 lines, chorus repeated max 3 times. Strongest line first in each section.
3. Vocal cues in (parentheses) where delivery matters.
4. A separate style prompt: 8-15 comma-separated tags for genre, mood, vocal direction, instruments, production/tempo. This goes in Suno's Style field, NOT the lyrics.

Rules:
- No prose directions inside the lyrics. Anything the singer should not sing goes in the style prompt.
- Keep it singable: short lines, real rhymes, no tongue-twisters.
- Show me the lyrics first. I will revise before generating.

Which chat tools work with this prompt?

Paste the prompt into any of these assistants (free or paid tiers).

• ChatGPT
• Claude

What tips make it smoother?

• Match the style prompt's mood to the lyrics' emotion. Mismatches produce confused songs.
• Change one variable at a time when testing: lyrics, style, or tags, never all three.
• Energy tags like [Build] and [Drop] shape the arrangement. Use them deliberately.
• Fewer than 5 style tags is too vague; more than 20 dilutes the direction.

Where do the claims come from?

Summaries use public sources linked below, not independent verification. How setups are picked.

• Read VoteMyAI’s write-up (www.votemyai.com)

  VoteMyAI: Section tag rules: Custom Mode only, Lyrics field only, one tag per line in brackets, never inline; the core and vocal-delivery tag sets.

• Read HookGenius’s write-up (hookgenius.app)

  HookGenius: Style prompt formula and tag counts; lyric best practices: verse/chorus line counts, chorus repetition limit, strongest line first.

• Read Jack Righteous’s write-up (jackrighteous.com)

  Jack Righteous: Field placement: styles in the Style field, words and section tags in Lyrics; prose in Lyrics can be sung; change one variable at a time.