Model reference · open weights
MiniMax-Music-3_GGUFs is an open-weight audio or speech model from realrebelai, listed in the AxForge catalogue. AxForge can bring it up on EU-owned hardware for you on request — with the licence handled where one is required.
About
MiniMax-Music-3 — GGUF GGUF quantizations of MiniMax-Music3, converted from Comfy-Org's fp32 repack for ComfyUI with ComfyUI-GGUF. MiniMax-Music-3 generates full music from a text prompt. The diffusion transformer is only 2.46B parameters — small enough that even the top tier fits comfortably on modest hardware, and Q4 runs on almost anything. Apache-2.0, inherited from the base model. Files Take a higher tier than you would on an image model. The fp32 source is 9.15 GB and Q80 is 2.6 GB — on a model this small there's little reason to drop low, and audio gives quantization error nowhere to hide. There's no visual channel to mask it: what would read as slightly soft detail in an image reads as smeared transients, ringing, or a thinner stereo field here. Q80 or Q6K is the sensible default. Companion files The GGUF replaces only the diffusion transformer. Both of these come from Comfy-Org/MiniMax-Music-3: A pruned text-encoder variant is also available in that repo if the bf16 encoder is too large for your setup. Usage 1. Put the .gguf in ComfyUI/models/diffusionmodels 2. Load it with Unet Loader (GGUF) 3. Text encoder, VAE and sampler settings are unchanged from the official MiniMax-Music-3 workflow Requires ComfyUI with MiniMax-Music-3 support, plus the ComfyUI-GGUF custom node. Conversion notes Converted from the official fp32 transformer with a streaming converter, quantized with llama-quantize, then repaired. This is a 1-D convolutional audio DiT, not a standard image-model architecture — preprocessconv [2304, 2304, 1], postprocessconv [128, 128, 1], latentconditioners [2048, 4096, 3] — with a separate timestep-feature path and a small condlayerlogits gate. Kept at high precision rather than quantized: The conv layers and the conditioning path are where a music model has the least tolerance for error — the timestep embedding steers every denoising step, and the latent conditioners carry the structure the audio is built on. llama-quantize does not preserve these automatically; they are restored afterward and verified per file. Every file was checked after conversion to confirm those layers are stored at F16/F32 and that the tensor count and architecture metada
Summarised from the published model card. Read the full card on the HuggingFace links below.
Specifications
| Maker | realrebelai |
|---|---|
| Type | Audio & music |
| Variants | 1 |
| Runs with | gguf |
| Based on | MiniMaxAI/MiniMax-Music3 |
| Released | 2026-08-14 |
| Popularity | 1k downloads / month |
| Likes | 8 |
| Licence | Open weights |
How it works
Variants
Open weights ship in several sizes and precisions. One page, all the variants — pick the one that fits your GPU. VRAM figures are estimates from model size.
| Variant | Params | Precision | VRAM | Fits 16 GB | Weights |
|---|---|---|---|---|---|
| MiniMax-Music-3_GGUFs | — | GGUF | — | — | Weights ↗ |
Using it via the API
Once AxForge deploys minimax-music-3-ggufs for you, it answers on the OpenAI-compatible API — the same base URL and keys as every other model. (minimax-music-3-ggufs below is illustrative; you get the exact model name on deployment.)
$ curl -sS https://api.axforge.ai/v1/audio/transcriptions \ -H "Authorization: Bearer $AXFORGE_API_KEY" \ -F model="minimax-music-3-ggufs" -F file=@audio.mp3
Details
Tags
Licence
Open weights under apache-2.0 — commercial use is permitted. Deploy it on AxForge EU hardware on request. Read the licence ↗