Generate Music with MiniMax Music 3

Create

Compose complete songs up to five minutes with MiniMax Music 3 — structured captions, sectioned lyrics, and 44.1 kHz stereo audio.

Generate music with MiniMax Music 3 — full songs up to five minutes

Describe a vibe, get a song.

MiniMax Music 3 is built for complete songs, not just short beds. Give it a structured style caption and sectioned lyrics — [verse], [chorus], [bridge] — and it returns a 44.1 kHz stereo track up to five minutes long. Use it for content soundtracks, scratch demos, mood-setting underscore, or starting points you'll re-record in a DAW. Leave lyrics blank or toggle instrumental for vocal-free beds. Every generation is async, polled by the orchestrator, and lands in your Tool Runs as a downloadable file.

How to generate music with MiniMax Music 3

Five steps from idea to a finished track.

1. Describe the music in concrete terms: genre, tempo (BPM), key, instrumentation, mood, and intended use.
2. Paste lyrics with structure tags on their own lines ([verse], [chorus], [bridge]) — or leave them blank for an instrumental.
3. Pick a duration. Credit cost scales with length so a one-minute bed stays cheaper than a five-minute song.
4. Submit; the model runs an async job and the orchestrator polls until the track is rendered.
5. Download or pipe the audio into a video edit; re-run with tweaks to genre or tempo to triangulate the feel.

What MiniMax Music 3 can score

Use-cases that pair well with a generated track.

Full songs

Up to five minutes

Write a verse-chorus structure and get a complete arrangement instead of a 30-second loop.

Background beds

Under-narration loops

Generate non-intrusive instrumental beds for VO-led content.

Pitch demos

Pre-clearance scratch

Mock up a song idea you can re-record cleanly with a band later.

Content soundtracks

Indie YT / TikTok

Score short-form video without the licensing rabbit hole.

Where MiniMax Music 3 fits in real workflows

Concrete use-cases that justify a dedicated landing page.

• Full-length song drafts with verse, chorus, and bridge
• Background beds for explainer and tutorial content
• Pitch demos to take into a studio session later
• Short-form soundtracks for Reels and TikTok content
• Branded podcast intros & outros without licensing friction

Why a dedicated MiniMax Music 3 workspace

And why pinning the model matters.

MiniMax Music 3 composes by genre, instrumentation, tempo, and sectioned lyrics — and it lives or dies by how concrete your prompt is. "A song that sounds happy" generates noise; "Bright synth-pop, 110 BPM, handclap-driven, 80s radio energy" generates a track. Put structure tags on their own lines. The orchestrator runs the model as an async job and stores the result so you can re-render variations cheaply.

Pro tips for MiniMax Music 3

Small adjustments that meaningfully improve output quality.

1. Use concrete musical language — tempo (BPM), genre, key, instruments — MiniMax Music 3 responds far better to "85 BPM lo-fi hip-hop" than to "chill vibes".
2. Put lyric structure tags on their own lines. Text on the same line as [chorus] is dropped.
3. Name the use case in the prompt ("under VO narration", "TikTok hook"); it nudges the model toward content-friendly mixes.
4. For instrumentals, toggle Is Instrumental or include "no vocals" in the prompt.
5. Start at 1 minute while you iterate, then stretch to 3–5 minutes once the arrangement is right.
6. Re-record in a DAW for production work. AI music is great for pitch demos and content soundtracks; less so for label releases.

MiniMax Music 3 on Gab AI — frequently asked questions

What is MiniMax Music 3?

MiniMax Music 3 is an AI music generator built by MiniMax. It creates complete songs up to five minutes from a style description and sectioned lyrics. On Gab AI it's available as a standalone, pinned tool — runs through the same orchestrator, credits, and file pipeline as chat.

Is this tool free to use?

Anyone with a Gab AI account can run MiniMax Music 3. Each run deducts the model's per-request credit cost from your balance — there's no surprise per-month fee.

What does it cost per run with MiniMax Music 3?

Credit cost is set on the underlying MiniMax Music 3 model and scales with duration. A one-minute track is 18 credits; longer songs add credits so five-minute generations stay profitable. The form shows the exact cost before you submit.

Can I use the output commercially?

Commercial-use rights depend on MiniMax's terms for MiniMax Music 3, not on Gab AI. AI-generated audio is generally usable in your own content; double-check the model card if you plan to release tracks for sale or sync.

Does MiniMax Music 3 clone artist voices or styles?

No. MiniMax Music 3 does not generate "in the style of" named artists in this pipeline. Refusals for copyrighted melodies and named-artist prompts are on the model layer.

Why a separate tool for every model?

Because every model is different and the multi-model picker quietly hides those differences. Pinning MiniMax Music 3 to its own tool gives you predictable cost, consistent style, and a fair lane for comparing one model's output against another's without confusing the cause of the difference.

Can I switch to a different model from here?

Yes — every model gets the same kind of landing page. Use the catalog at /tools to browse all model-playground tools, or pick a different one from the related tools section below.

Where do my runs go?

Every run lands in your Tool Runs (under My Library). You can revisit, download, fork, or continue any run in chat for follow-up work.

Ready to compose with MiniMax Music 3?

One model, one form, one good result.

Stop arguing with a model picker mid-project. Pin MiniMax Music 3 as your engine of choice, run the form above, and let the orchestrator handle credits, file storage, and run history exactly the way it does for chat. Everything you generate is yours, saved to your Tool Runs, and ready to fork or continue.