Keoo AI
Top AI Music & Audio Models

MiniMax: full AI songs and lifelike voices

MiniMax is an AI company with two standout audio tools. MiniMax Music 3.0 creates complete songs of up to five minutes, with vocals and full production. MiniMax Speech turns text into natural voices in more than 40 languages.

Explore MiniMax ↗
MiniMax creative workspace
Made by
MiniMax
Model type
AI music and voice models
Released
Music 3.0: summer 2026
Best for
Full 5-minute songs and lifelike voices

What is MiniMax?

MiniMax is a Chinese AI lab that also makes the Hailuo video models. Its Music 3.0 model arrived in summer 2026 and was later released with open weights. In one pass it writes, arranges, sings and produces a full song, with intro, verses, choruses, bridge and outro.

Music 3.0 understands detailed descriptions of how a song should change over time, such as tempo, key, which instruments come in and out, and how the vocals should feel. MiniMax Speech, its voice model, is used by many voice-agent companies and supports voice cloning across more than 40 languages.

What MiniMax does best

  • Songs up to 5 minutes. Complete structured tracks in one go.
  • Lyrics or instrumental. Bring your own words, or make music with no vocals.
  • Detailed control. Describe tempo, key, instruments and how the song builds.
  • Natural voices. MiniMax Speech sounds human in 40+ languages.
  • Commercial-friendly licence. MiniMax's community licence for Music 3.0 allows commercial projects with credit.

How to use MiniMax on Keoo AI

  1. Open the Music Generator (for songs) or Text to Speech (for voices) in Keoo AI.
  2. Pick MiniMax from the model list.
  3. For music: describe the style and paste lyrics. For voice: paste your script and choose a voice.
  4. Generate and download your audio.

What you can make with it

  • Full-length background music for videos and games
  • Song demos and sketches
  • Voice-overs for YouTube, adverts and courses
  • Multilingual narration
  • Voices for chatbots and phone agents

How to write a MiniMax prompt

Music: genretempokeyinstrumentshow it buildsvocal style. Voice: scriptvoicetone.
  • Describe how the song changes: 'starts with soft piano, drums enter at the chorus'.
  • Give a tempo (BPM) and a mood.
  • For voice, write the script the way people speak, with short sentences.
  • Use punctuation to control pauses in speech.

Example 1: YouTube intro voice (our sample)

Welcome back to the channel. Today I'm going to show you how to turn one simple idea into a week's worth of content, in under ten minutes. Let's get into it.

Example 2: Lo-fi study track

Lo-fi hip hop, 80 BPM, mellow and cosy. Starts with soft Rhodes piano and vinyl crackle, a gentle boom-bap beat comes in after 8 bars, warm bass, light rain in the background. Instrumental, no vocals.

Questions people ask

How long can MiniMax songs be?

Music 3.0 can create complete songs of up to five minutes.

Can MiniMax make voices too?

Yes. MiniMax Speech creates natural voices in more than 40 languages.

Can I use MiniMax music commercially?

MiniMax's community licence for Music 3.0 allows commercial use with credit. Check the licence for your project.

Create more with MiniMax on Keoo AI

Keoo AI puts MiniMax next to the other top video, image and audio models in one dashboard. Open Music Generator, pick your model, and go from idea to finished content in minutes.

Start creating for free ↗