LLM EXPLORER 62,589 MODELS INDEXED

DeepSeek V4.1 Flash Quark MXFP4 by amd

By amd · 660 downloads
HF Repo ↗ Copy URL

DeepSeek V4.1 Flash Quark MXFP4 is an open-source language model by amd. Features: 483.9b LLM, VRAM: 290.9GB, License: mit, LLM Explorer Score: 0.3.

  8-bit   Amd-quark Base model:deepseek-ai/deepsee... Base model:quantized:deepseek-...   Deepseek   Deepseek v41   Endpoints compatible   Moe   Ocp-mx   Quantization   Quark   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/amd/DeepSeek-V4.1-Flash-Quark-MXFP4 

DeepSeek V4.1 Flash Quark MXFP4 Parameters and Internals

LLM NameDeepSeek V4.1 Flash Quark MXFP4
Repository 🤗https://huggingface.co/amd/DeepSeek-V4.1-Flash-Quark-MXFP4 
Base Model(s)  deepseek-ai/DeepSeek-V4.1-Flash   deepseek-ai/DeepSeek-V4.1-Flash
Model Size483.9b
Required VRAM290.9 GB
Updated2026-09-30
Maintaineramd
Model Typedeepseek_v41
Model Files  1.0 GB: 1-of-48   1.3 GB: 2-of-48   7.4 GB: 3-of-48   7.4 GB: 4-of-48   7.4 GB: 5-of-48   7.4 GB: 6-of-48   7.4 GB: 7-of-48   7.4 GB: 8-of-48   7.4 GB: 9-of-48   7.4 GB: 10-of-48   7.4 GB: 11-of-48   7.4 GB: 12-of-48   7.4 GB: 13-of-48   7.4 GB: 14-of-48   7.4 GB: 15-of-48   7.4 GB: 16-of-48   7.4 GB: 17-of-48   7.4 GB: 18-of-48   7.4 GB: 19-of-48   7.4 GB: 20-of-48   7.4 GB: 21-of-48   7.4 GB: 22-of-48   7.4 GB: 23-of-48   7.4 GB: 24-of-48   7.4 GB: 25-of-48   7.4 GB: 26-of-48   7.4 GB: 27-of-48   7.4 GB: 28-of-48   7.4 GB: 29-of-48   7.4 GB: 30-of-48   7.4 GB: 31-of-48   7.4 GB: 32-of-48   7.4 GB: 33-of-48   7.4 GB: 34-of-48   7.4 GB: 35-of-48   7.4 GB: 36-of-48   7.4 GB: 37-of-48   7.4 GB: 38-of-48   7.4 GB: 39-of-48   7.4 GB: 40-of-48   7.4 GB: 41-of-48
Model ArchitectureDeepseekV41ForCausalLM
Licensemit
Transformers Version5.6.0