AI music model
MiniMax Music 2.0 — AI songs steered by your reference track
MiniMax Music 2.0 creates vocal tracks guided by an audio reference you upload. Use it when you already have a groove, vibe, or performance you want the new song to follow — then finish and download in Veontra.
What MiniMax Music 2.0 can do
Unlike text-only models, MiniMax Music 2.0 requires a reference audio file (MP3 or WAV longer than 15 seconds) that contains music — ideally with vocals. The model steers the new generation toward that sonic character while still following your style prompt and lyrics settings.
You can choose instrumental, auto lyrics, or custom lyrics. Track length is not set with a duration timer: it follows the lyrics and model behavior. That is why Veontra hides the duration selector for this model.
In the Veontra generator you still pick genre, mood, vocals, and tempo to shape the brief that wraps your prompt. The reference upload sits next to those controls so the workflow stays in one place.
Billing is a flat credit cost per generation — competitive when you want reference-driven vocals without paying per second of output.
How to generate with MiniMax Music 2.0 in Veontra
- Select MiniMax Music 2.0 and upload a clear reference clip (30–90 seconds of strong groove works well).
- Describe the song you want and set genre, mood, vocals, and tempo.
- Choose lyrics mode — instrumental, auto, or custom — and keep custom lyrics concise if you want a shorter track.
- Generate, listen in the feed, then download the MP3 from your workspace.
Who MiniMax Music 2.0 is for
Producers and songwriters who start from an existing vibe — a demo, a favorite arrangement, or a temp track — get the most from reference steering.
Content creators can match a campaign’s sonic brand by uploading a short house, pop, or hip-hop reference and generating a fresh vocal take that still feels related.
It is also useful for “style transfer” style experiments: keep the energy of a reference while rewriting lyrics and arrangement through your prompt.
How MiniMax Music 2.0 compares to other Veontra models
Pick MiniMax 2.0 when audio reference is required and you want vocal songs steered by that file. If you prefer image mood reference and explicit duration, choose Lyria 3 Pro. If you care most about lyrics-driven singing without uploading audio, choose MiniMax Music 2.6. For long instrumentals up to ten minutes, choose Sonilo v1.1.
Switch between models in the same picker — Learn more pages like this one explain each option before you spend credits.
MiniMax Music 2.0 FAQ
Is a reference track required?
Yes. MiniMax Music 2.0 will not generate without an audio reference (MP3/WAV, longer than 15 seconds).
Why is there no duration control?
The fal MiniMax 2.0 API does not take a duration parameter. Length follows lyrics and model defaults, so Veontra hides the timer for this model.
What makes a good reference?
Use a clean 30–90 second clip with music and ideally vocals, in the genre you want. Avoid spoken-word podcasts and very noisy recordings.
Can I write my own lyrics?
Yes — use custom lyrics mode, or auto lyrics if you only want a style prompt.
How is it billed?
Flat credits per generation in Veontra. Check Credit services for the current MiniMax Music 2.0 rate.
One AI workspace for your team
Extract documents, transcribe audio, and generate images from the same dashboard — with human review before anything is exported.