MiniMax Music - Kyma API

Overview

Part of the Music-2.0 family, this model converts text prompts up to 2000 tokens into audio tracks up to five minutes long. It operates in the strong quality tier and is designed specifically for lyrics-driven composition. Kyma serves MiniMax Music through an OpenAI-compatible endpoint with automatic request failover. Prompt caching is enabled, billing repeated prompt prefixes at this model’s cached input rate. Every response includes exact cost in usage.cost and reports the executing variant in the X-Kyma-Model header. The model is classified as slow and does not support structured outputs, vision, or real-time streaming. It is billed per song rather than by token, making it suitable for batch workflows but not interactive applications.

Specs

Field	Value
Model ID	`minimax-music`
Best for	Background music, social shorts, royalty-free soundtrack, bulk theme generation
Context window	2,000
Input modalities	Text
Output modalities	Audio
Tool calling	Yes
Structured outputs	Yes
Prompt caching	Yes
Speed	Slow
Cost band	Cheap
Release stage	Stable

Pricing

	Per 1M tokens
Input	$0.20
Output	$0.60

Use this when

Background Music Generation — Produce royalty-free audio tracks to accompany video content or podcasts.
Social Media Shorts Audio — Generate short, thematic soundtracks optimized for vertical video platforms.
Bulk Theme Generation — Create multiple musical variations at scale for prototyping or asset libraries.

Not ideal for

Do not use this model for real-time audio synthesis, interactive voice applications, or workflows requiring sub-second generation latency.

Pick something else when

You need faster audio generation: use eleven-turbo-v2-5.
You need higher-fidelity music tracks: use minimax-music-pro.
You need premium, hero-grade compositions: use elevenlabs-music.

Example

from openai import OpenAI

client = OpenAI(base_url="https://kymaapi.com/v1", api_key="ky-...")

response = client.chat.completions.create(
    model="minimax-music",
    messages=[{"role": "user", "content": "..."}],
)

Agent query example

Ask the API which models fit, instead of hardcoding an id:

curl "https://kymaapi.com/v1/models?recommended_for=generate-speech&tools=true&quality_tier=strong"

FAQ

How does Kyma handle billing for this model? Requests are billed per song rather than by token, and prompt caching bills repeated prompt prefixes at this model’s cached input rate. Does Kyma route requests automatically if a provider fails? Yes, every request includes automatic failover, and the exact model that ran is reported in the X-Kyma-Model response header. Can I use this model for real-time voice applications? No, the model is optimized for batch generation with a slow speed tier and does not support streaming or structured outputs.

​Overview

​Specs

​Pricing

​Use this when

​Not ideal for

​Pick something else when

​Example

​Agent query example

​FAQ