How to use Suno V5 metatags to control the structure of an AI-generated song. Verified list, examples, traps to avoid.
Metatags are the hidden control layer of Suno. Square-bracket markers like [Verse], [Chorus], [Whispered], [Drop] separate chaotic, rambling AI songs from structured, professional-sounding tracks. This is the verified list for Suno V5 / V5.5 (April 2026) — what is actually recognized, where to put it, and where it falls flat.
Metatags belong in the Lyrics field, never in Style. Two rules:
[Verse] on its own line, then the lyric on the next.Inline tags inside a lyric line are usually ignored. Empty trailing spaces around brackets sometimes break parsing — keep it tight.
[Verse 1]
I keep your photo in my coat
And read it like a note
[Pre-Chorus]
[Building]
But every time the song comes on
[Chorus]
[Belted]
I burn the letter and write againSuno reliably recognizes these section markers. The "reliable" qualifier matters: the model treats metatags as signal weights, not hard switches. They bias section identity, energy and arrangement density. They do not force exact production outcomes.
| Tag | Effect |
|---|---|
[Intro] | Opening — short, often sparse. Vocals optional. |
[Verse] / [Verse 1] / [Verse 2] | Verse. Numbered variants help variation across verses. |
[Pre-Chorus] | Lift before chorus — typically rising melody, building energy. |
[Chorus] | Hook section — biggest energy, most repetition. |
[Post-Chorus] | Tail after chorus — vocal hooks, often wordless ("oohs"). |
[Bridge] | Mid-song break, contrast. |
[Build] / [Rise] | Rising build-up — risers, drum rolls, filter sweeps. |
[Drop] | Energy peak / impact — best paired with [Build] right before. |
[Break] / [Breakdown] | Pause / dropout — strips arrangement back. |
[Instrumental] | No vocals on this section. Pair with solo tags below. |
[Guitar Solo] / [Drum Solo] / [Saxophone Solo] | Specific instrumental section. Suno honors these well. |
[Outro] / [Ending] | Closing — often spare, fading. |
[Fade Out] / [Fade In] | Volume envelope cues. Reliable. |
Voice tags shape the vocal performance for the section that follows. They are most effective when placed on a line right after the section tag and before the lyric:
| Tag | Effect |
|---|---|
[Whispered] | Intimate, breathy delivery. |
[Belted] | Power-vocal projection. |
[Falsetto] | Head voice, light register. |
[Spoken] / [Spoken Word] | Talked, not sung. |
[Rap] | Rap delivery. Pair with verse tag. |
[Choir] | Multi-voice arrangement. |
[Harmonized] | Vocal stack / multi-part harmony. |
[Auto-Tuned] | Pitch-corrected, processed. |
[Ad-lib] | Improvised vocal layer / runs. |
[A Cappella] | Strips backing — vocals only. |
[Hook] | Earworm phrase, often repeated. |
[Chant] | Group / repetitive vocal pattern. |
These are softer cues — they describe the feel of the section, not the structure. Use sparingly.
| Tag | Effect |
|---|---|
[Soft] / [Gentle] | Low intensity. |
[Loud] / [Aggressive] | High intensity. |
[Upbeat] / [Mellow] | Tempo / mood signal. |
[Explosive] | Maximum impact — useful before drops. |
[Intimate] | Close-mic, sparse. |
[Epic] | Cinematic scale — risk of over-bombast. |
[Minimal] / [Stripped Back] | Removes arrangement layers. |
[Layered] | Adds vocal/instrumental layers. |
Cues that mark a change between sections. Best placed on a line of their own between two section tags.
| Tag | Effect |
|---|---|
[Riser] | Synthetic rising sound — lifts into the next section. |
[Impact] | Hit / boom — punctuates a transition. |
[Reverse] | Reverse cymbal / reversed sound. |
[Stutter] | Stutter edit — chopped repetition. |
[Tape Stop] | Signature pitch-bend pause. |
[Filter Sweep] | Filter movement across the bar. |
[Key Change] | Modulation. Suno honors this well. |
[Half-Time] / [Double-Time] | Tempo feel switch within the same BPM. |
The most-used metatag combo in dance music. Stack them like this:
[Verse]
We're losing time tonight
[Build]
[Riser]
And every second feels too tight
[Drop]
[Instrumental]Suno will create risers, drum rolls and a low-frequency build under the Build tag, then a high-energy impact at the Drop. The pattern is well-supported in EDM, drum & bass, dubstep, and big-room contexts.
The Style field is genre + mood + sound. Structure tags there are ignored at best, confusing at worst.
Two [Chorus] tags within the same 30-second window will produce a generic loop. Space your structure tags — the model reads pacing.
And then [Whispered] I told her is parsed unreliably. Put [Whispered] on its own line first, then the lyric.
[Sad], [Heartbroken], [Triumphant] are weak signals. Express emotion through the lyric, the chord movement and the voice tag — not through metatags.
[Verse 1] + [Whispered] + [Soft] + [Intimate] + [Vintage] dilutes the signal. Keep it to 2 tags max per section: one structural, one performance.
Specifically: Style (genre, instruments, mood), BPM, vocal gender generally, and language. All of those go in the Style field. Metatags are for shape and performance, not for sound identity.
For the full Suno prompt formula, see our complete Suno guide. To see how the metatag ecosystem maps across multiple platforms, the ToneDraft music board exposes the same vocabulary in a curated UI — pick from a verified tag list, and copy a clean Suno-ready output.
If you remember nothing else, remember these:
[Intro]
[Verse]
[Pre-Chorus]
[Chorus]
[Bridge]
[Outro]
[Build]
[Drop]
[Whispered]
[Belted]
[Falsetto]
[Instrumental]That covers 95% of usable cases. The rest is fine-tuning.