Downloads · 30 days
123
3% of all-time downloads
Minimax-M3-v0-NVFP4-REAP50
sparkarena/Minimax-M3-v0-NVFP4-REAP50
Minimax-M3-v0-NVFP4-REAP50 is a image-text-to-text model from sparkarena. Use it for the image-text-to-text task on the model card, and read the license before you ship it in a product. It is set up for transformers. The card lists the license as other.
This is an experimental quantization of MiniMax M3 to NVFP4 for use on DGX Spark.
All-time downloads
4.4K
Public
Parameters
125B
129 GB on disk
Likes
14
Public
Repo makeup
Click a slice to open those files.
.safetensors129 GB · 100%
Parameter types
How the weights are stored.
U8109B · 87%
Try a prompt
Base models
Model card
From the Hugging Face model README
NOTICE
This is an experimental quantization of MiniMax M3 to NVFP4 for use on DGX Spark.
On top of that, REAP 50% to make it fix on 2x DGX Sparks instead of 4x.
NVFP4 uses w1/3 scales, so need either: https://github.com/scitrera/sglang/tree/nvfp4-w13-scale-normalization (or possibly sglang PR#27588) in place to properly use the calibration from this model.
The calibration is still a work in progress (hence "v0" model). Updates are planned to improve performance.
Re: REAP50, I've tested that it's coherent, but I haven't exhaustively tested the quality degradation associated with REAP50. YMMV.
Run with sparkrun; part of Spark Arena
https://sparkrun.dev https://spark-arena.com
To run with sparkrun on 2x DGX Spark Nodes:
sparkrun run @experimental/minimax-m3-v0-nvfp4-2x-reap50
MiniMax-M3 is a native multimodal model with 1M context. It has ~428B parameters and ~23B activated parameters.
Highlights:
- Native Multimodality: M3 undergoes mixed-modality training from the very first step, enabling deeper semantic fusion across text, image, and video.
- Context Scaling via Sparse Attention: M3 introduces MiniMax Sparse Attention (MSA) to improve long context efficiency. M3 delivers 9× prefill and 15× decode speedups compared to M2 at 1M context, reducing per-token compute to 1/20.
- Coding & Cowork Capability: M3 achieves frontier-level performance across long-horizon agentic benchmarks, excelling in both coding and cowork.
MiniMax Sparse Attention (MSA)
M3 is powered by MiniMax Sparse Attention (MSA), a high-performance sparse attention operator designed for million-token contexts. Compared with GQA, MSA dramatically reduces the attention compute and memory footprint while preserving model quality.
<p align="center"> <img width="100%" src="figures/efficiency_gqa_vs_msa.png" alt="GQA vs MSA Efficiency Comparison"> </p>📄 Read the technical report: arXiv:2606.13392 · Hugging Face Papers
How to Use
M3 supports two reasoning modes:
- thinking — for complex reasoning, agentic tasks, and long-horizon collaboration.
- non-thinking — for latency-sensitive scenarios such as chat and code completion.
Local Deployment
Download the model:
hf download MiniMaxAI/MiniMax-M3 --local-dir MiniMax-M3
We recommend the following inference frameworks (listed alphabetically) to serve the model:
-
SGLang - see SGLang cookbook.
-
vLLM - see vLLM recipes.
-
Transformers - see Transformers docs.
Inference Parameters
We recommend the following parameters for best performance: temperature=1.0, top_p=0.95, top_k=40.
Contact Us
Contact us at [email protected].
- Task
- Image-Text-to-Text
- Library
- transformers
- Type
- minimax_m3_vl
- License
- other
- Created
- Jun 12, 2026
- Updated
- Jun 15, 2026