Does the lyrics generator produce audio?

No — it outputs text only. That is deliberate: it is a preprocessing step. Feed the lyrics into DiffRhythm or MiniMax Music to render actual audio.

What lyric format do the music models need?

DiffRhythm accepts either LRC format with timestamps, or section markers with at least two sections. The lyrics generator produces the sectioned format directly.

Can I use my own lyrics instead?

Yes. Skip the lyrics model and feed your own text straight into a music model, as long as it carries section structure.

AI Lyrics Generator

Structured lyrics with verses and a chorus — then turn them into a finished song.

Start creating

What is an AI lyrics generator?

An AI lyrics generator writes a complete song from a theme or style description — title, verses, chorus, structure — rather than a loose block of text. Structure is the part that matters downstream: music models need lyrics marked into sections before they can render a track, so unstructured output means doing that work by hand. On Floniks the lyrics model is designed as a preprocessing step, and its output feeds straight into a music model in the same workflow.

Models behind this tool

Every model listed here is already available on your account. Credit cost per render is read live from the model catalogue.

MiniMax Lyrics5 credits
DiffRhythm (FAL)50 credits
MiniMax Music 2.675 credits
Lyria 2 (FAL)50 credits

What people use it for

Original songs from a theme

Go from a one-line brief to a structured lyric sheet ready for rendering.

Content soundtracks

Write lyrics that name a product or campaign, then render a track around them.

Songwriting drafts

Produce several structural options quickly, then rewrite the lines worth keeping.

How it works

Describe the theme and style
Genre, mood and subject. The more specific the brief, the less generic the chorus.
Generate structured lyrics
Output arrives with a title and section markers, which is the format music models expect.
Render the track
Feed the lyrics into DiffRhythm for a full song, or MiniMax Music for ambient and background tracks.

Frequently asked questions

Does the lyrics generator produce audio?: No — it outputs text only. That is deliberate: it is a preprocessing step. Feed the lyrics into DiffRhythm or MiniMax Music to render actual audio.
What lyric format do the music models need?: DiffRhythm accepts either LRC format with timestamps, or section markers with at least two sections. The lyrics generator produces the sectioned format directly.
Can I use my own lyrics instead?: Yes. Skip the lyrics model and feed your own text straight into a music model, as long as it carries section structure.