NVIDIA-Nemotron-Parse-2.0

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 token, and updated training coverage for chart/table-heavy documents. NVIDIA Nemotron Parse 2.0 is intended for document understanding, information retrieval, data extraction, and multimodal data-curation workflows.

This model is ready for commercial or non-commercial use.

License/Terms of Use:

This model and its associated configuration files are licensed under the OpenMDW License Agreement, version 1.1 (OpenMDW-1.1). Use of the tokenizer included in this model is governed by the CC-BY-4.0 license.

This project will download and install additional third-party open source software projects. Review the license terms of these open source projects before use. Contributions are accepted under the policy in CONTRIBUTING.md.

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 can be used to convert 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.

Capability Highlights:

• Expanded multilingual OCR support, with substantial gains on CJK and Indic-script document text.
• Improved handwritten-text extraction for document pages containing informal, handwritten, or note-like content.
• Chart-to-table parsing that can identify chart regions and convert visible chart information into structured text for downstream use.
• Improved table handling, including stronger table detection, structure recovery, and text extraction on table-heavy documents.

Release Date:

Hugging Face August 3, 2026 on the NVIDIA Nemotron Parse 2.0 model page

References(s):

• Hugging Face Transformers mBART documentation
• NVIDIA C-RADIO

Model Architecture:

Architecture Type: Transformer-based vision-encoder-decoder model

Network Architecture:

• Vision Encoder: ViT-H model based on NVIDIA C-RADIO
• Adapter Layer: 1D convolutions and normalization layers that compress the vision latent sequence before decoding
• Decoder: mBART decoder with 10 blocks
• Auxiliary Prediction Head: One training-time decoder prediction head is preserved separately in auxiliary_prediction_heads.safetensors.extra for future multi-token prediction research. Standard generation uses the tied decoder input/output embeddings; the default model.safetensors, Transformers examples, and vLLM examples do not load this auxiliary head.
• Tokenizer: The tokenizer contains 72,256 entries, including an approximately 20k-token expansion over NVIDIA Nemotron Parse v1.2 to improve multilingual token efficiency. The model also includes task/control tokens such as , , , and , plus the chart class token ``. Use of the tokenizer included in this model is governed by the CC-BY-4.0 license.
• Number of Parameters:

Input(s):

Input Type(s): Image, Text

Input Format(s):

• Image: Red, Green, Blue (RGB)
• Text: Prompt string

Input Parameters:

• Image: Two-Dimensional (2D)
• Text: One-Dimensional (1D)

Other Properties Related to Input:

• Recommended maximum input resolution (Width, Height): 1664, 2048
• Recommended minimum input resolution (Width, Height): 1024, 1280
• Channel Count: 3
• Prompt format: one task prompt composed from the supported control tokens. The default prompt extracts bounding boxes, semantic classes, and text in markdown format: . The model can emit chart regions using when chart content is detected.

Output(s):

Output Type(s): Text

Output Format(s): String

Output Parameters: One-Dimensional (1D)

Other Properties Related to Output: Nemotron Parse 2.0 returns a string encoding document text, semantic element classes, and bounding boxes. Postprocessing utilities in this repository can transform generated bounding boxes back to original image coordinates and convert table or chart-associated text into LaTeX, HTML, markdown, JSON, or CSV where supported.

Our AI models are designed and/or optimized to run on NVIDIA GPU-accelerated systems. By leveraging NVIDIA hardware and software frameworks, the model achieves faster training and inference times compared to CPU-only solutions.

Quick Start

Direct Transformers inference dependencies

This installation is only for the direct Transformers example in the next section. It is not needed for the vLLM container workflow below. You can use the public image nvcr.io/nvidia/pytorch:25.03-py3 with the following library versions installed on top:

pip install accelerate==1.12.0
pip install transformers==5.6.1
pip install timm==1.0.22
pip install open_clip_torch==3.2.0
pip install einops==0.8.1
pip install beautifulsoup4

open_clip_torch is currently needed only by the direct Transformers path because C-RADIO's remote-code validation inspects an optional OpenCLIP adaptor. Nemotro