MiniMax-Music needs about 2.1 GB of memory at Q4_K (a 1.4 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 13 quants, from 0.1 GB to 17.2 GB (BF16).
MiniMax-Music 3 GGUF – 4.9B parameter music generation model
MiniMax-Music 3 is a 4.9B parameter model that generates full songs up to five minutes. It uses a hierarchical architecture with an 8B global LLM and a 0.6B local LLM, producing 32 kHz, 16‑bit stereo WAV audio. The GGUF package supports multiple quantizations for performance tuning.
Good at
- Long‑form song generation up to five minutes
- Fine‑grained control via lyrics and music description
- High‑quality 32 kHz stereo audio output
Limits
- No vision support
- Requires 8 k context for estimates
- License is other – check compatibility
Best for
- Music producers needing full‑song generation
- Researchers testing hierarchical audio models
- Developers building music‑generation tools
Who it’s for. This model is suitable for developers and researchers who need a high‑performance music generation engine that can produce long, coherent songs with detailed control over style and structure. It is also useful for audio‑engineering projects that require high‑fidelity 32 kHz stereo output.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 0.1 GB | 0.6 GB | Fits | Get file → | |
| — | 0.2 GB | 0.7 GB | Fits | Get file → | |
| Q4_K | 0.4 GB | 1.0 GB | Fits | Get file → | |
| Q8_0 | 0.7 GB | 1.3 GB | Fits | Get file → | |
| BF16 | 1.3 GB | 1.9 GB | Fits | Get file → | |
| Q4_0 | 1.4 GB | 2.1 GB | Fits | Get file → | |
| Q4_K | 1.4 GB | 2.1 GB | Fits | Get file → | |
| Q8_0 | 2.6 GB | 3.4 GB | Fits | Get file → | |
| Q4_0 | 6.0 GB | 7.2 GB | Tight | Get file → | |
| Q4_K | 7.2 GB | 8.5 GB | Too big | Get file → | |
| BF16 | 9.7 GB | 11.4 GB | Too big | Get file → | |
| Q8_0 | 9.9 GB | 11.6 GB | Too big | Get file → | |
| BF16 | 17.2 GB | 19.7 GB | Too big | Get file → |
For 8 GB VRAM, use the Q4_0 or Q4_K GGUFs (~1.4 GB file, ~2.1 GB memory). For 16 GB VRAM, Q8_0 (~2.61 GB file, ~3.4 GB memory) or BF16 (~1.29 GB file, ~1.9 GB memory) provide higher quality. Larger VRAM allows the 9.89 GB Q8_0 or 9.73 GB BF16 variants for maximum fidelity.
What GPU do you need for MiniMax-Music?
These figures add the file size, an fp16 KV cache at an 8K context window and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: the best quant that fits is Q8_0 (2.6 GB file, about 3.4 GB in use).
- 6 GB VRAM: the best quant that fits is Q8_0 (2.6 GB file, about 3.4 GB in use).
- 8 GB VRAM: the best quant that fits is Q4_0 (6.0 GB file, about 7.2 GB in use).
- 12 GB VRAM: the best quant that fits is Q8_0 (9.9 GB file, about 11.6 GB in use).
- 16 GB VRAM: the best quant that fits is Q8_0 (9.9 GB file, about 11.6 GB in use).
- 24 GB VRAM: the best quant that fits is BF16 (17.2 GB file, about 19.7 GB in use).
Which MiniMax-Music quant should you choose?
Q4_K (1.4 GB) is the middle option in this repository. The usual sweet spot between file size and quality, and the quant most people start with.
If memory is tight, the smallest file is the first one in the table at 0.1 GB, with the quality trade-offs that come with fewer bits.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about MiniMax-Music
How much VRAM does MiniMax-Music need?
About 2.1 GB at Q4_K with an 8K context window. That is the 1.4 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run MiniMax-Music on an 8 GB GPU?
Yes, with Q4_0: the 6.0 GB file needs about 7.2 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of MiniMax-Music GGUF?
It depends on the quant: file is 0.1 GB, file is 0.2 GB, Q4_K is 0.4 GB, Q8_0 is 0.7 GB, BF16 is 1.3 GB, Q4_0 is 1.4 GB and 7 more.
Where can I download MiniMax-Music GGUF?
The files are on Hugging Face in the audio-cpp/MiniMax-Music3-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does MiniMax-Music use?
Hugging Face lists the license as other. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run MiniMax-Music 3 on 8 GB VRAM?
Yes, the Q4_0 or Q4_K GGUFs use about 2.1 GB memory at 8 k context, fitting within 8 GB VRAM.
Does the model support image input?
No, MiniMax-Music 3 does not support vision or image inputs.
What is the maximum song length the model can generate?
The model can generate songs up to five minutes long with full structural coherence.