LLM EXPLORER 61,836 MODELS INDEXED

DeepSeek V4 Pro NVFP4 by nvidia

By nvidia · 73114 downloads
HF Repo ↗ Copy URL

DeepSeek V4 Pro NVFP4 is an open-source language model by nvidia. Features: 910b LLM, VRAM: 575.2GB, Context: 1024K, License: mit, LLM Explorer Score: 0.34.

  8-bit Base model:deepseek-ai/deepsee... Base model:quantized:deepseek-...   Deepseek v4   Deepseekv4   Fp8   Model optimizer   Modelopt   Nvfp4   Nvidia   Quantized   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/nvidia/DeepSeek-V4-Pro-NVFP4 

DeepSeek V4 Pro NVFP4 Parameters and Internals

LLM NameDeepSeek V4 Pro NVFP4
Repository 🤗https://huggingface.co/nvidia/DeepSeek-V4-Pro-NVFP4 
Base Model(s)  DeepSeek V4 Pro   deepseek-ai/DeepSeek-V4-Pro
Model Size910b
Required VRAM575.2 GB
Updated2026-08-07
Maintainernvidia
Model Typedeepseek_v4
Model Files  1.9 GB: 1-of-64   14.7 GB: 2-of-64   14.7 GB: 3-of-64   14.7 GB: 4-of-64   14.7 GB: 5-of-64   14.7 GB: 6-of-64   14.7 GB: 7-of-64   14.7 GB: 8-of-64   14.7 GB: 9-of-64   14.7 GB: 10-of-64   14.7 GB: 11-of-64   14.7 GB: 12-of-64   14.7 GB: 13-of-64   14.7 GB: 14-of-64   14.7 GB: 15-of-64   14.7 GB: 16-of-64   14.7 GB: 17-of-64   14.7 GB: 18-of-64   14.7 GB: 19-of-64   14.7 GB: 20-of-64   14.7 GB: 21-of-64   14.7 GB: 22-of-64   14.7 GB: 23-of-64   14.7 GB: 24-of-64   14.7 GB: 25-of-64   14.7 GB: 26-of-64   14.7 GB: 27-of-64   14.7 GB: 28-of-64   14.7 GB: 29-of-64   14.7 GB: 30-of-64   14.7 GB: 31-of-64   14.7 GB: 32-of-64   14.7 GB: 33-of-64   14.7 GB: 34-of-64   14.7 GB: 35-of-64   14.7 GB: 36-of-64   14.7 GB: 37-of-64   14.7 GB: 38-of-64   14.7 GB: 39-of-64   14.7 GB: 40-of-64
Model ArchitectureDeepseekV4ForCausalLM
Licensemit
Context Length1048576
Model Max Length1048576
Transformers Version4.57.1
Vocabulary Size129280
Torch Data Typebfloat16