Paper page - NVIDIA Nemotron Nano V2 VL (original) (raw)
Published on Nov 6, 2025
·
Submitted by 
taesiri
on Nov 7, 2025
Authors:
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
Abstract
Nemotron Nano V2 VL, a hybrid Mamba-Transformer LLM, improves document and video understanding through enhanced architecture and token reduction techniques.
We introduce Nemotron Nano V2 VL, the latest model of the Nemotron vision-language series designed for strong real-world document understanding, long video comprehension, and reasoning tasks. Nemotron Nano V2 VL delivers significant improvements over our previous model, Llama-3.1-Nemotron-Nano-VL-8B, across all vision and text domains through major enhancements in model architecture, datasets, and training recipes. Nemotron Nano V2 VL builds on Nemotron Nano V2, a hybrid Mamba-Transformer LLM, and innovative token reduction techniques to achieve higher inference throughput in long document and video scenarios. We are releasing model checkpoints in BF16,FP8, and FP4 formats and sharing large parts of our datasets, recipes and training code.
View arXiv page View PDF GitHub 1.62k auto Add to collection
Get this paper in your agent:
hf papers read 2511.03929
Don't have the latest CLI?
curl -LsSf https://hf.co/cli/install.sh | bash
Models citing this paper 4
nvidia/NVIDIA-Nemotron-Nano-12B-v2-VL-NVFP4-QAD Image-Text-to-Text • 8B • Updated May 5 • 3.32k • 27
richtext/NVIDIA-Nemotron-Nano-12B-v2-VL-BF16 Image-Text-to-Text • 13B • Updated Apr 2 • 1
Datasets citing this paper 2
nvidia/Nemotron-VLM-Dataset-v2 Viewer • Updated Dec 18, 2025• 4.58M • 6.22k • 91
introvoyz041/Nemotron-VLM-Dataset-v2 Viewer • Updated Dec 25, 2025• 4.58M • 1.38k