LLM EXPLORER 65,695 MODELS INDEXED

MiniMax M3.1 Flash Pruned 272B NVFP4 by primitive-ai

By primitive-ai · 104 downloads
HF Repo ↗ Copy URL

MiniMax M3.1 Flash Pruned 272B NVFP4 is an open-source language model by primitive-ai. Features: 272b LLM, VRAM: 137GB, License: other.

  8-bit   Agentic Base model:minimaxai/minimax-m... Base model:quantized:minimaxai...   Compressed-tensors   Conversational   Custom code   En   Expert-pruning   Image-text-to-text   Minimax-m3   Minimax m3 vl   Moe   Mxfp8   Nvfp4   Reasoning   Region:us   Safetensors   Sharded   Tensorflow   Tool-calling   Vllm
Model Card on HF 🤗: https://huggingface.co/primitive-ai/MiniMax-M3.1-Flash-pruned-272B-NVFP4 

MiniMax M3.1 Flash Pruned 272B NVFP4 Parameters and Internals

LLM NameMiniMax M3.1 Flash Pruned 272B NVFP4
Repository 🤗https://huggingface.co/primitive-ai/MiniMax-M3.1-Flash-pruned-272B-NVFP4 
Base Model(s)  MiniMaxAI/MiniMax-M3.1-Flash   MiniMaxAI/MiniMax-M3.1-Flash
Model Size272b
Required VRAM137 GB
Updated2026-10-27
Maintainerprimitive-ai
Model Typeminimax_m3_vl
Model Files  8.7 GB: 1-of-48   2.7 GB: 2-of-48   2.7 GB: 3-of-48   5.5 GB: 4-of-48   5.5 GB: 5-of-48   5.5 GB: 6-of-48   5.5 GB: 7-of-48   5.5 GB: 8-of-48   5.5 GB: 9-of-48   5.5 GB: 10-of-48   5.5 GB: 11-of-48   5.5 GB: 12-of-48   3.2 GB: 13-of-48   2.7 GB: 14-of-48   2.7 GB: 15-of-48   2.7 GB: 16-of-48   2.7 GB: 17-of-48   2.7 GB: 18-of-48   2.7 GB: 19-of-48   2.7 GB: 20-of-48   2.7 GB: 21-of-48   2.7 GB: 22-of-48   2.7 GB: 23-of-48   2.7 GB: 24-of-48   2.7 GB: 25-of-48   2.7 GB: 26-of-48   2.7 GB: 27-of-48   2.7 GB: 28-of-48   2.7 GB: 29-of-48   2.7 GB: 30-of-48   2.7 GB: 31-of-48   2.7 GB: 32-of-48   2.7 GB: 33-of-48   2.7 GB: 34-of-48   2.7 GB: 35-of-48   2.7 GB: 36-of-48   2.7 GB: 37-of-48   2.7 GB: 38-of-48   2.7 GB: 39-of-48
Supported Languagesen
Model ArchitectureMiniMaxM3SparseForConditionalGeneration
Licenseother
Transformers Version4.52.4
Torch Data Typebfloat16