Lyric formats, explained: Suno tags, lyric sheets and timed lyrics

Lyric sheets and timed lyrics

Here’s the simplest way to think about it. There are really only two kinds of lyric format. A lyric sheet is the words, usually with a few labels saying which part is the verse and which part is the chorus. A timed lyric file is the words plus when they’re sung, either line by line or, if it’s fancy, word by word.

You write sheets. Music players, captions and lyric videos need timing. Getting from one to the other is the tricky bit, because somebody (or something) has to sit with the song and mark the exact moment every line begins. If you’ve ever tried doing that by hand for a whole song, you know how long it takes.

Lyric sheet formats

Plain text

The simplest lyric sheet is just the words, with each sung line on its own line and a blank line between stanzas. Nothing fancy. It’s also exactly what Apple Music asks artists for. Their guidelines say “all lyric lines must be single-spaced and a double space must separate each stanza.” They also want no commas or periods at the end of a line, every repeat written out and background vocals in parentheses (Apple Music for Artists).

Left the porch light on for nobody
Kept your jacket on the chair (On the chair)

So call me when the city sleeps
Call me when it’s late

Section tags in square brackets: Genius and Suno

You’ve almost certainly seen this one. Genius labels each section in square brackets on its own line, like [Verse 1], [Chorus] and [Bridge], and when more than one artist is on a track, it adds who’s on that part: [Verse 2: Artist Name].

AI music apps picked up the same habit. Suno’s own guide tells you to “use structure tags like [Verse], [Chorus]” (Suno).

[Verse 1]
Left the porch light on for nobody
Kept your jacket on the chair (On the chair)

[Chorus]
So call me when the city sleeps
Call me when it’s late

Section tags kept apart: Spotify and Musixmatch

Spotify does things a little differently. In most countries its lyrics come from Musixmatch, and the guidelines ask you to “label sections with tags like #VERSE and #CHORUS.” Those tags label the sections, but listeners never see them as lyrics. Spotify also asks you to write every repeated line out in full, and not to “use multipliers like ‘x2’” (Spotify for Artists). Keep that last rule in mind. It’s going to come up again.

ChordPro

If you play guitar, you might know ChordPro. It puts the chords right in the line, in square brackets, like [G]Left the porch light [D]on, and wraps a chorus in {start_of_chorus} and {end_of_chorus} (ChordPro). It’s great for musicians. For anything else, you’ll want to take the chords out first.

Timed lyric formats

LRC

LRC is the simplest timed format there is. Each line gets a timestamp in square brackets in front of it, and you can add a few optional tags at the top for the title, the artist and a timing offset. Music players and karaoke apps read it, and YouTube will take it for captions too (YouTube Help). There’s also an “enhanced” version that puts a timestamp in front of every word, which is what you need if you want words to light up one at a time.

[ti:Porch Light]
[ar:Your Name]
[00:12.40]Left the porch light on for nobody
[00:16.85]Kept your jacket on the chair
[00:21.10]<00:21.10>So <00:21.42>call <00:21.80>me <00:22.05>when

One small catch: LRC uses square brackets for its own timestamps and tags, so labels like [Chorus] are best left out of LRC files.

SRT and WebVTT

These two are caption formats, the files you upload next to a video on YouTube. Each caption has a start time, an end time and a line or two of text. WebVTT can go a step further and time the individual words inside a caption.

1
00:00:12,400 --> 00:00:16,300
Left the porch light on for nobody

2
00:00:16,850 --> 00:00:20,600
Kept your jacket on the chair

TTML for Apple Music

Apple Music’s synced lyrics use something called TTML. It’s an XML format, and it can time every line, every word, even every syllable. It also knows which part of the song you’re in, which singer has each line, and which words are background vocals, so they can be shown apart from the lead (Apple Video and Audio Asset Guide). It’s the most detailed format here. It’s also the one you’re least likely to ever write by hand, and honestly, that’s fine.

The formats side by side

Whatever the format, the people who publish lyrics for a living agree on three simple rules. Write every repeat out instead of “x2.” Put backing vocals and ad-libs in parentheses. And start a new line whenever the music or the delivery changes.

What Suno does with lyrics

When Suno writes the lyrics

If you describe a song in Simple mode, or ask Suno to write lyrics for you in Advanced mode (it used to be called Custom), its lyric writer “drafts in song structure with verse, chorus, pre-chorus, and bridge labels” (Suno).

In the public songs we looked through, that means bracket tags sitting on their own lines: [Verse 1], [Pre-Chorus], [Chorus], [Bridge], [Outro], and every so often [Final Chorus] or [End]. For Spanish or Italian songs the tags are often in the song’s own language, like [Estribillo] or [Ritornello]. Choruses get written out every time, and echoes go in parentheses. Here’s roughly what that looks like:

[Verse 1]
Left the porch light on for nobody
Kept your jacket on the chair (On the chair)

[Pre-Chorus]
And I know, I know

[Chorus]
So call me when the city sleeps
Call me when it’s late

When you write your own

Here’s the good news if you write your own lyrics. Whatever you type into Advanced mode is saved just the way you typed it. Suno doesn’t sneak in tags or reformat anything, so if you write without tags, your song’s lyrics won’t have any. And when you copy the lyrics from a song page, the tags and blank lines come right along with them.

It helps to think of tags as suggestions, not commands. Suno’s own guides only name a handful, like [Verse], [Chorus] and [Bridge]. Those huge lists of tags for vocals, instruments and moods floating around online? They come from other people’s experiments, not from Suno. As for parentheses, most people use them for backing vocals and echoes, and that’s how Suno’s own lyric writer uses them too.

What the song actually sings

Now for the part that catches people out. Suno never actually promises to sing exactly what’s in the lyrics box.

In our experience, after thousands of generations on the newer models, it usually does. When it doesn’t, the differences tend to be small: an extra ad-lib here, an outro or chorus repeated one more time there, and every now and then a tag sung out loud as if it were part of the song. Big rewrites are rare. They do happen, though.

Older models wandered off a bit more. A 2025 research paper, written by the team behind a rival model, found Suno v4.5 “redundantly repeating the chorus at the end” (SongBloom).

So it’s worth remembering that the lyrics on a Suno song page tell you what was asked for, which isn’t always what was sung. Suno doesn’t give you a timed lyrics file either. You can download an MP3 on every plan, a WAV on Pro and Premier, a video, stems, and MIDI for songs made in Suno Studio (Suno Help), but nothing that tells you when each line happens.

Lines, phrases and bars

People use three words for lyrics pretty interchangeably, and it gets confusing. Here’s how they actually differ:

• A bar (or measure) is a chunk of musical time, usually four beats in hip-hop and pop. Rappers count their verses in bars, so a 16-bar verse is sixteen of those chunks.
• A phrase is a musical sentence. It’s often two or four bars long, and in a sung melody it’s roughly one breath.
• A line is simply what you write down.

Here’s the thing: no lyric standard says one line has to equal one bar. The people who publish lyrics start a new line when the music or the delivery changes, like a new section, a change in tempo or a switch from singing to rapping. In practice, rappers tend to write one bar per line, singers tend to write one phrase per line, and both read naturally.

The format we recommend

If you’re making a lyric video, the best lyric sheet is a clean one. Everything that’s sung, nothing that isn’t, laid out the way it’s actually performed. Here’s how we’d write it:

1. Each sung line on its own line. Hit Enter wherever the singer moves on to a new line: after each bar if you’re rapping, after each breath if you’re singing.
2. A blank line between sections.
3. Section tags on their own lines, like [Chorus] or #CHORUS.
4. Every repeat written out. Never “x2.”
5. Backing vocals and ad-libs in parentheses, closed on the same line.
6. Only what’s sung. Leave out performer names, stage directions like [Whisper] or [Guitar Solo], translations and links.
7. The lyrics as sung, not as written. If the vocal says it differently, change the sheet to match.

Put together, it looks like this:

[Verse 1]
Left the porch light on for nobody
Kept your jacket on the chair (On the chair)
Every song the radio plays
Sounds like you were there

[Chorus]
So call me when the city sleeps
Call me when it’s late
I’ll be the one still wide awake
Call me anyway (Call me anyway)

[Verse 2]
Took the long way past your building
Counted windows, lost my place

[Chorus]
So call me when the city sleeps
Call me when it’s late
I’ll be the one still wide awake
Call me anyway (Call me anyway)

From a Suno song to a lyric video

Here’s the whole trip, start to finish:

1. Download your song from Suno as an MP3, or a WAV if your plan includes it.
2. Copy the lyrics from the song’s page. Tags and all is fine.
3. Start a new song in thesis, add the audio, paste in the lyrics and pick English or Spanish.
4. Let thesis time it. It separates the vocal and times every word for you.
5. Listen along and fix anything Suno sang a little differently.
6. Make your video. Download a lyric video with SRT captions, or let the AI Director make one in any of seventeen looks.

Want to see more? Take a look at our page on music videos for Suno songs.

Questions