NVIDIA Nemotron Parse 2.0 Overview

Description:

NVIDIA Nemotron Parse 2.0 transforms document images into structured, machine-readable representations with text, layout classes, bounding boxes, and reading-order information. Given a Red, Green, Blue (RGB) document image and a task prompt, the model produces formatted text and spatial annotations for document elements such as titles, paragraphs, captions, tables, charts, page headers, page footers, footnotes, pictures, and bibliography entries. Compared with NVIDIA Nemotron Parse v1.2, NVIDIA Nemotron Parse 2.0 adds an approximately 20k-token vocabulary expansion for more efficient multilingual support, chart-aware document parsing with the <class_Chart> class token, and updated training coverage for chart/table-heavy documents and stronger support for many languages. NVIDIA Nemotron Parse 2.0 is intended for document understanding, information retrieval, data extraction, and multimodal data-curation workflows. NVIDIA Nemotron Parse 2.0 was developed by NVIDIA as a part of Nemotron.
This model is ready for commercial or non-commercial use.

License/Terms of Use:

GOVERNING TERMS: Use of this trial service is governed by the NVIDIA API Trial Terms of Service. Use of the model is governed by the OpenMDW License Agreement, version 1.1.

Deployment Geography:

Global

Use Case:

NVIDIA Nemotron Parse 2.0 is designed for developers and teams building document intelligence, retrieval-augmented generation (RAG), curator, extractor, and agentic AI applications. It converts scanned or rendered PDFs, presentation slides, forms, reports, tables, and mixed-content document pages into structured outputs for downstream indexing, retrieval, analytics, model training-data creation, and human-in-the-loop review. The model supports document understanding, OCR, layout parsing, table extraction, and text extraction with spatial grounding.

Release Date:

Build.NVIDIA.com: 09/11/2026 via link HuggingFace: 06/30/2026 via link

Reference(s):

Hugging Face Transformers mBART documentation
NVIDIA C-RADIO

Model Architecture:

Architecture Type: Transformer
Network Architecture: Vision Encoder (ViT-H based on NVIDIA C-RADIO) + Adapter (1D convolutions and normalization layers) + Decoder (mBART with 10 blocks)
This model was developed based on mBART.
Number of model parameters: 905M

Computational Load

Cumulative Compute: 1.50x10^21 FLOP
Estimated Energy and Emissions for Model Training: ~4.64 MWh

Input:

Input Type(s): Image, Text
Input Format(s):

• Image: Red, Green, Blue (RGB)
• Text: String

Input Parameters: Two-Dimensional (2D), One-Dimensional (1D)
Other Properties Related to Input: The model accepts RGB images with a recommended maximum resolution of 1664×2048 pixels and a minimum of 1024×1280 pixels. Input prompts consist of control tokens; the default prompt is </s><output_markdown> to extract bounding boxes, classes, and markdown text. The model can also emit <class_Chart> tokens for detected chart regions.

Output:

Output Type(s): Text
Output Format: String
Output Parameters: One-Dimensional (1D)
Other Properties Related to Output: Maximum output length configurable via max_new_tokens parameter (default 9000, max determined by generation config). Supports streaming output via transformers generate. Output is UTF-8 encoded text in formats such as markdown, LaTeX, HTML, JSON, or plain text.

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):

• Transformers
• vLLM

Supported Hardware Microarchitecture Compatibility:

• NVIDIA Ampere
• NVIDIA Blackwell
• NVIDIA Hopper
• NVIDIA Turing

Supported Operating System(s): Linux

The integration of foundation and fine-tuned models into AI systems requires additional testing using use-case-specific data to ensure safe and effective deployment. Following the V-model methodology, iterative testing and validation at both unit and system levels are essential to mitigate risks, meet technical and functional requirements, and ensure compliance with safety and ethical standards before deployment.

This AI model can be embedded as an Application Programming Interface (API) call into the software environment described above.

Model Version(s):

NVIDIA Nemotron Parse 2.0

Integration requires Transformers or vLLM frameworks and runs on Linux with NVIDIA Ampere, NVIDIA Blackwell, NVIDIA Hopper, or NVIDIA Turing GPUs.

Training, Testing, and Evaluation Datasets:

Training Dataset:

Data Modality:

• Image
• Text

Image Training Data Size: 1 Million to 1 Billion Images
Text Training Data Size: 1 Billion to 10 Trillion Tokens
Data Collection Method by dataset: Hybrid: Automated/Manually-Collected/Synthetic
Labeling Method by dataset: Hybrid: Automated/Manually-Labelled/Synthetic
Properties (Quantity, Dataset Descriptions, Sensor(s)): The training set contains millions of image-text items aggregated from large document, table, and layout datasets. It consists of document-page and table images paired with OCR text, bounding boxes, and layout labels. Sources include rendered digital documents, scientific papers, PDFs, Wikipedia-style pages, and synthetic document, table, word, and character renderings. Annotations are derived from OCR and layout models, third-party OCR services, synthetic-generation pipelines, and human labeling.

Testing Dataset:

Data Collection Method by dataset: Hybrid: Automated/Manually-Collected/Synthetic
Labeling Method by dataset: Hybrid: Automated/Manually-Labelled/Synthetic
Properties (Quantity, Dataset Descriptions, Sensor(s)): Tested on internal and public document-understanding benchmarks covering OCR quality, layout structure, table parsing, reading order, and visual grounding.

Evaluation Dataset:

Benchmark Score: See benchmark table below.

Benchmark coverage:

• ParseBench evaluates document parsing across text fidelity, semantic formatting, tables, charts, and visual grounding with layout boxes/classes.
• IndicVisionBench evaluates OCR quality for Indic-language pages, including Bengali, Gujarati, Hindi, Kannada, Malayalam, Marathi, Odia, Punjabi, Tamil, and Telugu.
• MOSCAR evaluates multilingual synthetic OCR across broad script coverage, including Latin, Arabic, Cyrillic, Chinese, Hangul, Japanese, Indic scripts, Hebrew, Thai, Greek, and others.
• OmniDocBench Notes (Handwriting) evaluates text-block edit distance on note-style document pages from the data_source: note slice.

The following results compare NVIDIA Nemotron Parse 2.0 with NVIDIA Nemotron Parse v1.2 on internal and public evaluation benchmarks. NVIDIA Nemotron Parse 2.0 results use the equal-weight checkpoint soup from training checkpoints 58k, 60k, 62k, 64k, and 66k. Unless marked otherwise, metrics are higher-is-better; arrows indicate the direction of the 2.0 change relative to v1.2.

Data Collection Method by dataset: Hybrid: Automated/Manually-Collected
Labeling Method by dataset: Hybrid: Automated/Manually-Labelled
Properties (Quantity, Dataset Descriptions, Sensor(s)): Evaluated on internal and public document-understanding benchmarks covering OCR quality, layout structure, table parsing, reading order, and visual grounding.

Inference:

Acceleration Engine: vLLM Test Hardware:

• NVIDIA Ampere (A100, A10G)
• NVIDIA Blackwell (B200, GB200, RTX PRO 6000)
• NVIDIA Hopper (H100)
• NVIDIA Lovelace (L40S)

Ethical Considerations:

NVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. Developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse.
Please make sure you have proper rights and permissions for all input image and video content; if image or video includes people, personal health information, or intellectual property, the image or video generated will not blur or maintain proportions of image subjects included.
For more detailed information on ethical considerations for this model, please see the Model Card++ Explainability, Bias, Safety & Security, and Privacy Subcards.

Please report model quality, risk, security vulnerabilities or NVIDIA AI Concerns here.

Generated by NVIDIA Model Card Generator Toolkit.