LLM EXPLORER 62,155 MODELS INDEXED

MiniMax M3 NVFP4 by nvidia

By nvidia · 403798 downloads
HF Repo ↗ Copy URL

MiniMax M3 NVFP4 is an open-source language model by nvidia. Features: 246.6b LLM, VRAM: 51.9GB, License: other, LLM Explorer Score: 0.38.

  8-bit Base model:minimaxai/minimax-m... Base model:quantized:minimaxai...   Conversational   Custom code   Minimax-m3   Minimax m3 vl   Modelopt   Nvfp4   Nvidia   Quantized   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/nvidia/MiniMax-M3-NVFP4 

MiniMax M3 NVFP4 Parameters and Internals

LLM NameMiniMax M3 NVFP4
Repository 🤗https://huggingface.co/nvidia/MiniMax-M3-NVFP4 
Base Model(s)  MiniMax M3   MiniMaxAI/MiniMax-M3
Model Size246.6b
Required VRAM51.9 GB
Updated2026-07-23
Maintainernvidia
Model Typeminimax_m3_vl
Model Files  6.0 GB: 1-of-88   0.4 GB: 2-of-88   0.4 GB: 3-of-88   0.4 GB: 4-of-88   0.4 GB: 5-of-88   0.4 GB: 6-of-88   0.4 GB: 7-of-88   0.4 GB: 8-of-88   0.4 GB: 9-of-88   0.4 GB: 10-of-88   0.4 GB: 11-of-88   0.4 GB: 12-of-88   0.4 GB: 13-of-88   0.4 GB: 14-of-88   0.4 GB: 15-of-88   0.4 GB: 16-of-88   0.4 GB: 17-of-88   0.4 GB: 18-of-88   0.4 GB: 19-of-88   0.4 GB: 20-of-88   0.4 GB: 21-of-88   0.4 GB: 22-of-88   0.4 GB: 23-of-88   0.4 GB: 24-of-88   0.4 GB: 25-of-88   0.4 GB: 26-of-88   0.4 GB: 27-of-88   0.2 GB: 28-of-88   2.1 GB: 29-of-88   0.2 GB: 30-of-88   0.2 GB: 31-of-88   4.1 GB: 32-of-88   4.1 GB: 33-of-88   4.1 GB: 34-of-88   4.1 GB: 35-of-88   4.1 GB: 36-of-88   4.1 GB: 37-of-88   4.1 GB: 38-of-88   4.1 GB: 39-of-88
Model ArchitectureMiniMaxM3SparseForConditionalGeneration
Licenseother
Transformers Version4.52.4
Torch Data Typebfloat16