---
tags:
- blackhole
- p150
- tt-dit-server
- tt-model-cache
- tt-model-catalog
- tt-model-container
---

# minimax-music3

MiniMax Music 3, a lyrics-and-caption conditioned music generation model (8B Qwen3-based global LLM, 0.6B depth decoder for the residual RVQ codebooks, 2.4B flow-matching DiT and a DAC-style vocoder; 44.1 kHz stereo, served at 32 kHz like the reference server) running on a single Tenstorrent Blackhole chip. SGLang-Omni compatible /v1/audio/speech API plus an async jobs API for multi-minute songs. Model weights under the MiniMax-Music3 Community License (products must display "MiniMax-Music3").

Runs on **p150** (mesh `P150`).

Packaged and published with [tt-model-manager](https://github.com/tenstorrent/tt-model-manager) 0.1.0 (manifest schema 5.1).

## Quickstart

```bash
tt-model pull  jashansinghTT/minimax-music3-blackhole --with-weights
tt-model serve jashansinghTT/minimax-music3-blackhole
```

`pull --with-weights` downloads the Docker image and the [`MiniMaxAI/MiniMax-Music3`](https://huggingface.co/MiniMaxAI/MiniMax-Music3) weights at `fbdf52fbaaca799592917417eb05f1899f1255ec` (into your HF cache; they are not in the image). `serve` starts the model's own HTTP server on port 20000 (or the next free port, if that one is busy); the first start compiles kernels for your device, which takes several minutes, and the server is ready when it logs `Application startup complete`.

### Try it
```bash
curl -s localhost:20000/v1/health
curl -s localhost:20000/v1/audio/speech -H 'Content-Type: application/json' -d '{
  "model": "MiniMaxAI/MiniMax-Music3",
  "input": "[Verse]\nMorning light filtering through the pine\n[Chorus]\nSoftly the world begins to breathe",
  "instructions": "A warm acoustic pop song with intimate female vocals, fingerpicked guitar and soft piano.",
  "response_format": "wav", "seed": 7, "max_new_tokens": 750}' -o song.wav
```
`max_new_tokens` counts audio frames (25 per second). Long songs: `POST /v1/music/jobs` then poll `GET /v1/music/jobs/{id}`
and fetch `GET /v1/music/jobs/{id}/audio`.

## Provenance

The exact sources the image was built from — `code/` in this repo is byte-identical to the model code inside the image:

| component | built from |
| --- | --- |
| tt-metal | a local checkout — commit not published |
| `code/` digest | `9a9c37114f86ca58` (sha256, first 16 hex digits) |
| built | 2026-09-12T09:31:21+00:00 by tt-model 0.1.0 |