NVIDIA-Nemotron-Nano-VL

License/Terms of Use

Governing Terms:

Your use of the model is governed by the NVIDIA Open License Agreement.

Additional Information:

Backbone LLM: NVIDIA-Nemotron-Nano-12B-v2.

Deployment Geography:

Global

Use Case:

Customers: AI foundry enterprise customers

Use Cases: Image summarization. Text-image analysis, Optical Character Recognition, Interactive Q&A on images, Text Chain-of-Thought reasoning

Release Date:

• Build.Nvidia.com [October 28th, 2025] via nvidia/NVIDIA-Nemotron-Nano-VL-12B-V2
• Hugging Face [October 28th, 2025] via nvidia/NVIDIA-Nemotron-Nano-VL-12B-V2-BF16
• Hugging Face [October 28th, 2025] via nvidia/NVIDIA-Nemotron-Nano-VL-12B-V2-FP8
• Hugging Face [October 28th, 2025] via nvidia/NVIDIA-Nemotron-Nano-VL-12B-V2-NVFP4

Model Architecture:

Network Type: Transformer

Network Architecture:

Vision Encoder: C-RADIOv2-H

Language Encoder: NVIDIA-Nemotron-Nano-12B-v2

Input

Input Type(s): Image, Text

• Input Images
• Language Supported: German, Spanish, French, Italian, Korean, Portuguese, Russian, Japanese, Chinese, English

Input Format(s): Image (Red, Green, Blue (RGB)), and Text (String)

Input Parameters: Image (2D), Text (1D)

Other Properties Related to Input:

• Context length up to 128K
• Maximum Resolution: Determined by a 12-tile layout constraint, with each tile being 512 × 512 pixels. This supports aspect ratios such as:
  • 4 × 3 layout: up to 2048 × 1536 pixels
  • 3 × 4 layout: up to 1536 × 2048 pixels
  • 2 × 6 layout: up to 1024 × 3072 pixels
  • 6 × 2 layout: up to 3072 × 1024 pixels
  • Other configurations allowed, provided total tiles ≤ 12
• Channel Count: 3 channels (RGB)
• Alpha Channel: Not supported (no transparency)

Output

Output Type(s): Text

Output Formats: String

Output Parameters: One-Dimensional (1D): Sequences up to 128K

Our AI models are designed and/or optimized to run on NVIDIA GPU-accelerated systems. By leveraging NVIDIA’s hardware (e.g. GPU cores) and software frameworks (e.g., CUDA libraries), the model achieves faster training and inference times compared to CPU-only solutions.

Software Integration

Runtime Engine(s): vLLM Supported Hardware Microarchitecture Compatibility: H100 SXM 80GB Supported Operating System(s): Linux

Model Versions:

Nemotron-Nano-VL-12B-V2-FP8

Quick Start

Install Dependencies

pip install causal_conv1d "transformers>4.53,<4.54" torch timm "mamba-ssm==2.2.5" accelerate open_clip_torch numpy pillow

Usage

To serve this checkpoint with vLLM, you can start the docker vllm/vllm-openai:nightly and run the sample command below:

python3 -m vllm.entrypoints.openai.api_server --model nvidia/Nemotron-Nano-VL-12B-V2-FP8 --trust-remote-code --quantization modelopt

Training, Testing, and Evaluation Datasets:

Training Datasets:

Data Modalities ** Total Size: 39'486'703 samples ** Total Number of Datasets: 270 ** Text-only datasets: 33 ** Text-and-image datasets: 176 ** Video-and-text datasets: 61 ** Total size: 27.7 TB

** Data modalities: Text, Image, Video ** Data Collection Method by dataset: Hybrid: Automated, Human, Synthetic ** Labeling Method by dataset: Hybrid: Automated, Human, Synthetic

** Dataset partition: Training [100%], Testing [0%], Validation [0%] ** Time period for training data collection: 2023-2025 ** Time period for testing data collection: N/A ** Time period for validation data collection: N/A

The post-training datasets consist of a mix of internal and public datasets designed for training vision language models across various tasks. It includes:

• Public datasets sourced from publicly available images and annotations, supporting tasks like classification, captioning, visual question answering, conversation modeling, document analysis and text/image reasoning.
• Internal text and image datasets built with public commercial images and internal labels, adapted for the same tasks as listed above.
• Synthetic image datasets generated programmatically for specific tasks like tabular data understanding and optical character recognition (OCR), for English, Chinese as well as other languages.
• Video datasets supporting video question answering and reasoning tasks from publicly available video sources, with either publicly available or internally generated annotations.
• Specialized datasets for safety alignment, function calling, and domain-specific tasks (e.g., science diagrams, financial question answering).
• NVIDIA-Sourced Synthetic Datasets for text reasoning.
• Private datasets for safety alignment or VQA on invoices.
• Crawled or scraped captioning, VQA, and video datasets.
• Some datasets were improved with Qwen2.5-72B-Instruct annotations

For around ~30% of our total training corpus and several of the domains listed above, we used commercially permissive models to perform:

• Language translation
• Re-labeling of annotations for text, image and video datasets
• S