From lyrics to a finished song in LUVI: MiniMax Lyrics, then Music 3.0 or Suno v6
Write the words with MiniMax Lyrics Generation (20 credits), then turn them into a song with MiniMax Music 3.0 (300 credits, one track up to five minutes) or Suno v6 in custom mode (264 credits for two takes). The lyric sheet's section tags carry straight over.
A song with words takes two models in LUVI, and both arrived on October 17, 2026. MiniMax Lyrics Generation writes the lyric sheet for 20 credits. MiniMax Music 3.0 sings it for 300 credits, or Suno v6 does in custom mode for 264 credits, which buys two takes. Prices are flat per generation on all three: song length and settings don't change them. This post is about the hand-off between the two steps, because that is where the words get lost or kept.
What does MiniMax Lyrics Generation give you?
MiniMax Lyrics Generation returns text, not audio: a title, a style line and a lyric sheet divided into sections such as [Verse] and [Chorus]. In LUVI the result is stored as one text file laid out in that order: the title, a line that starts with Style:, a blank line, then the lyrics. It opens as text you can copy.
The model works in two modes, and the Lyrics field decides which one runs:
- Write. Leave the Lyrics field empty and put a one-line theme in the prompt ("a late train home after a double shift"). An empty prompt is allowed too; you get a song on a random subject.
- Edit or continue. Paste lyrics you already have into the Lyrics field and the prompt becomes the instruction: "make the chorus sadder", "add a second verse". If you fill in the Title field, the result keeps that title unchanged.
The prompt takes up to 2,000 characters and the Lyrics field up to 3,500. MiniMax's own guide describes this as the first of two steps: "first, use the Lyrics Generation API to write lyrics based on a theme; then, feed those lyrics into the Music Generation API to compose and produce a complete song."
How do the lyrics go into MiniMax Music 3.0?
MiniMax Music 3.0 splits the job between two fields: the prompt carries the sound and the Lyrics field carries the words. Paste the lyric sheet into Lyrics and write the style as the prompt. The Style: line from the lyric sheet is a good starting point for that prompt; add instruments, tempo and the kind of voice you want.
The section tags carry over without edits. Music 3.0 reads tags on their own line to shape the arrangement: [Intro], [Verse], [Pre Chorus], [Chorus], [Post Chorus], [Hook], [Bridge], [Interlude], [Transition], [Build Up], [Break], [Inst], [Solo] and [Outro]. A blank line is a pause.
Three behaviors are worth knowing before you spend 300 credits:
- An empty Lyrics field is not an instrumental. Music 3.0 starts with that field empty, and when it stays empty the model writes lyrics itself from your prompt and sings them. You only need the Lyrics step when the words matter to you.
is_instrumentalignores the lyrics. Switch it on and whatever is in the Lyrics field is not sung.- One track per generation, up to five minutes, as MP3 or WAV, at up to 44.1 kHz. The bitrate setting (up to 256 kbps) applies to MP3 only.
How do they go into Suno v6 instead?
Suno v6 takes lyrics only in custom mode. Switch custom on and the prompt itself is sung: paste the lyric sheet into the prompt (up to 3,000 characters), the style line into style, and the title into title. With custom off, the prompt is read as a description of the song, and Suno's custom-mode fields (title, style, negative_tags, duration and the two weighting sliders) are ignored without a warning.
What you gain is two takes for 264 credits, and a duration target from 15 seconds to six minutes that only works in custom mode. In LUVI's onboarding test on October 16, 2026, a 30-second target on Suno v6 mini came back as two tracks of 29.89 and 29.93 seconds. If you want Suno to rework your lines rather than sing them as written, auto_lyrics allows that, again in custom mode only. All three Suno v6 engines on the Suno Chirp family page take these same settings at the same price.
Which singer should get the words?
Pick by what you need to hand over at the end, not by the 36 credits between them:
- MiniMax Music 3.0 when the lyric sheet is final and its structure matters: the full tag set, up to five minutes, WAV output for an edit. Lyrics plus song costs 320 credits.
- Suno v6 when you want to hear two readings of the same words and choose, or need the song to land near a set length. Lyrics plus song costs 284 credits.
Either way, read the lyric sheet before you send it on. The lyric step costs 20 credits; a song built on a line you didn't want costs 264 or 300.
How we checked
- Models: MiniMax Lyrics Generation, MiniMax Music 3.0, Suno v6 and Suno v6 mini.
- What was read, and when: On October 20, 2026, the parameter contract of each model through the LUVI connector, live credit quotes for each with several settings, the catalog price on the model pages, and LUVI's onboarding record for these models, including its test run of October 16, 2026 (one generation per case).
- Results: 20 credits for Lyrics Generation, 300 for Music 3.0 with MP3 or WAV, and 264 for Suno v6 with any length target: none of the settings changed a price. In the onboarding run, Music 3.0 with an empty Lyrics field wrote and sang its own words over a 206-second track, and the lyric step returned a title, a style line and tagged sections. No outputs were generated for this post, so it makes no claim about output quality.
- Drawbacks: We did not compare how faithfully each singer follows a given lyric sheet, or how either handles lyrics outside English.
In LUVI
Open MiniMax Lyrics Generation in the workspace, write a theme, and copy the sheet when it lands. Then open MiniMax Music 3.0 or Suno v6 and paste it where that model expects words. If you work from Claude through the LUVI connector, the connector returns the lyric sheet as text, so Claude can read it and pass it on to the singing step in the same conversation.
New accounts start with 200 credits, and verifying a phone and a card adds up to 1,500 more. You can create a LUVI account and write your first verse before you pay for a note.
Sources
- Music Generation guide — MiniMax API Docs, developer guide. Accessed Oct 20, 2026.
- Lyrics Generation API reference — MiniMax API Docs. Accessed Oct 20, 2026.
- Introducing v6 — Suno, official blog, Sep 9, 2026. Accessed Oct 20, 2026.
More from Workflows
- A music bed for your clip: Suno v6 or MiniMax Music, and the switch that keeps it wordlessSuno v6 returns two takes for 264 credits; MiniMax Music 2.6 and 3.0 return one track for 300, with WAV output. For a bed under a clip, the setting that matters is Instrumental only: MiniMax switches it on when your prompt says instrumental, Suno v6 does not.
- Your model stops at ten seconds. Four ways to keep going, and a 10× price spreadFour models on LUVI extend a clip you already have. On the same ten-second source, adding five seconds quoted 450 credits on one and 4,510 on another. The craft rule is the same on all four: write only what happens next.
- Dubbing a clip into another language on LUVI: four steps and what each one costsTranscribe, rewrite, voice, re-sync the mouth. Every step is a separate model, and the last one decides the bill: two lip-sync models quoted 130 and 2,200 credits for the same ten-second clip.