01
Style plus lyrics
Combine a production-oriented prompt with your own lyric sheet so composition and vocal delivery are anchored to the same brief.
MiniMax Music 3.0 available in CAVN
MiniMax Music AI music generatorDirect the lyrics, sound, and output quality
MiniMax Music is strongest when you already know what the song should say and how it should sound. In CAVN, you can pair a detailed style prompt with your lyrics, title, and output settings in one generation request.
3.0
model used for this sample
44.1k
maximum available sample rate
256k
maximum available bitrate
Generated with MiniMax Music
Real CAVN API output · music-3.0
CAVN generated sample
Neon Tide
MP3 / 44.1 kHz / 256 kbps
Our test, not a stock demo
The exact brief behind Neon Tide
Generated on CAVN with MiniMax Music 3.0 using original English lyrics, a detailed style prompt, 44.1 kHz sample rate, and 256 kbps output bitrate.
Generation direction
“Cinematic synth-pop with pulsing analog bass, glassy arpeggios, crisp electronic drums, intimate female vocal, a widescreen chorus, and a nocturnal but hopeful mix.”
Custom lyric songsProduction-specific prompts44.1 kHz deliverables
What you can control
MiniMax Music capabilities inside CAVN
Choose MiniMax Music for directed song production: explicit lyrics, a specific sonic brief, and selectable audio quality. It gives less lyric improvisation than a pure prompt workflow and more control over the material you provide.
02
Instrumental mode
Generate without vocals when the job is background music, a cue, a topline foundation, or an arrangement starting point.
03
Audio settings
Select supported sample rates up to 44.1 kHz and bitrates up to 256 kbps for a predictable output target.
04
Structured request
Keep the title, lyrics, production prompt, and technical output settings in separate fields for a clearer, repeatable generation brief.
Model comparison
Compare Mureka, MiniMax Music, ACE-Step, and HappyShrimp
The useful question is not which model wins every prompt. It is which input style matches the material you already have and the decisions you need to control.
| Decision | Mureka | MiniMax Music | ACE-Step | HappyShrimp |
|---|---|---|---|---|
| Best starting point | A natural-language song idea | Finished lyrics plus a production brief | A brief with musical constraints | A one-line idea or reference audio |
| Primary control | Prompt expansion and variants | Lyrics, style, and audio settings | Lyrics, BPM, key, meter, and language | Lyrics mode, reference audio, and advanced settings |
| Current model | Mureka V9 | Music 3.0 | ACE-Step 1.5 Turbo | Web-based AI music generator |
| Ideal workflow | Explore several song directions quickly | Turn prepared writing into a directed song | Translate a music brief into explicit controls | Start from a prompt in the official product |
A practical MiniMax workflow
Keep the first generation measurable: one clear brief, one model, and a small number of candidates. Change only the inputs that need improvement on the next pass.
Open the music generator- 01
Prepare the song material
Write the lyrics and title, then describe instrumentation, vocal character, genre, energy, mix, and emotional arc.
- 02
Set output quality
Choose MiniMax Music 3.0 and the supported sample rate and bitrate that match your review or publishing workflow.
- 03
Generate and evaluate
Review lyric adherence, arrangement, vocal delivery, and mix translation before selecting a candidate for the next stage.
Know before generating
Limits worth planning around
The style prompt and lyric sheet are separate inputs. Keep production language in the prompt and singable words in the lyrics field.
The current CAVN MiniMax endpoint exposes sample-rate options up to 44.1 kHz and bitrate options up to 256 kbps.
A clean technical export does not replace creative review, mastering, or a rights check before release.
FAQ
Questions about MiniMax Music
What is MiniMax Music 3.0?+
MiniMax Music 3.0 is a music generation model available in CAVN. It creates a song from a style prompt and can also use supplied lyrics, a title, instrumental mode, and audio quality settings.
Can MiniMax Music use my own lyrics?+
Yes. The CAVN MiniMax workflow accepts lyrics separately from the style prompt, which makes it suitable for songs where the words are already written.
Can MiniMax Music generate instrumental tracks?+
Yes. Set instrumental mode when you want music without a generated vocal performance.
What quality is the MiniMax demo on this page?+
Neon Tide was generated with MiniMax Music 3.0 as a 44.1 kHz, 256 kbps MP3 using the prompt shown on this page and original English lyrics.
Explore other music models
Compare with another real sample
Your next track starts here
Ready to create with MiniMax?
Open the CAVN music generator with MiniMax Music selected, add your own brief, and turn it into a track you can refine.