What is Suno Lyrics Generator?
Suno Lyrics Generator writes song lyrics from a short description and returns them as text. Unlike the other Suno operations, it does not need a clip_id or any audio: the whole request is a prompt that describes the song and an mv value that picks the lyrics model, remi-v1 or default. Every completed job returns two options, labelled a and b. Each one has the lyrics in a text field with section tags like [Verse] and [Chorus], a suggested title, and a list of style tags such as "rock" or "electric guitar". Those fields line up with what Suno music generation expects in Custom mode, so the usual flow is to generate lyrics here, pick or edit one option, then send it on to render audio.
Key features of Suno Lyrics Generator
Three features make this endpoint useful on its own.
Two options per request
Every job returns a and b. You get a built-in choice without sending the prompt twice, which suits apps that let users pick between drafts.
Title and style tags included
Each option comes with a title and a tags list next to the lyrics. That gives you a song name and a style starting point for the music request in one call.
Text only, no audio rendering
The input is just prompt and mv, and the output is text. Use it to settle on the words before spending time on audio generations.
Best for
Lyric step before music generation
Generate lyrics, let a person edit them, then send the final text to Suno music in Custom mode.
Songwriting apps
Show users the two returned options side by side and let them choose or combine lines.
Getting past a blank page
Get two takes on a theme as a starting point for your own writing.
Batch content
Write lyrics for many short songs, jingles or greetings from a list of prompts, then render only the ones you approve.
Use cases
A song-creation app can split its flow into two screens: the first calls this endpoint and shows options a and b, the second sends the chosen lyrics, title and tags to the Suno music generation API. A greeting-card service can generate personalised lyrics from a short form ("birthday song for Anna, upbeat pop") and only render audio once the customer approves the words. A music teaching tool can generate two lyric sets on the same theme and ask students to compare their structure and rhyme. Teams that already have a melody can send the chosen lyrics to the Suno Add Vocals API to have them sung over an uploaded instrumental.
API examples
The full reference, including the completed response shape, is at docs.unifically.com/models/audio/suno/lyrics.
curl -X POST https://api.unifically.com/v1/tasks \
-H "Content-Type: application/json" \
-H "Authorization: Bearer YOUR_API_KEY" \
-d '{
"model": "suno-ai/lyrics",
"input": {
"prompt": "A chill rock song about summer days at the beach",
"mv": "remi-v1"
}
}'
The call returns a task_id. Poll GET /v1/tasks/<task_id> until the status is completed and read the two options from output, or set a callback_url to receive them.
FAQs
People also ask
Text, not audio. Each completed job returns two lyric options under output.a and output.b. Each option has a text field with the lyrics (section tags such as [Verse] included), a suggested title, and a list of tags describing the style, such as "chill rock".
The endpoint always returns a pair so you can pick the better one, merge lines from both, or show both to a user as choices.
Describe the song in plain words, including genre, mood and subject. "A chill rock song about summer days at the beach" is enough. prompt is the only content field.
They are the two lyrics models you can pick with mv. remi-v1 is the default in the playground; default is the alternative. Try both on the same prompt and keep the one whose writing you prefer.
Paste the chosen text into the prompt field of a Suno music generation request in Custom mode, and use the returned tags as a starting point for the tags field.
