nvidia/Nemotron-3.5-Parse
image-text-to-text model by nvidia
Older snapshot: check the original source for current information.
From the publisher
Model card & documentation
Source previewRead the publisher’s intended use, setup instructions, evaluations and limitations. The original model card is the source of truth.
Nemotron 3.5 Parse is a document image parser that produces structured, machine-readable text with semantic layout classes, bounding boxes, and reading order. It combines a C-RADIOv4-SO400M vision encoder with a 10-layer mBART decoder that consumes the visual sequence as a causal prefix. The release uses dynamic rectangular NaViT preprocessing, content-normalized coordinates, grouped-query self-attention (16 query heads and 8 key/value heads), a 72,505-token multilingual tokenizer, and chart-aware output through the token. It is intended for document understanding, information retrieval, data extraction, and multimodal data-curation workflows. Nemotron 3.5 Parse was developed by NVIDIA as a part of Nemotron. This model is ready for commercial or non-commercial use. Detailed benchmarks: ParseBench · IndicVisionBench · MOSCAR. ParseBench results use triple pass with region union on 2,078 unique pages. See pipeline details, reproduction instructions, and evaluation results. Additional OCR highlights (scores on a 0–100 scale): MOSCAR uses simple synthetic renderings of text in each language…
Read the full model card ↗ · Preview checked 2026-10-27T22:19:52.773Z
Inside the original model card — Document outline
- Nemotron 3.5 Parse
- Description:
- Benchmark highlights
- License/Terms of Use:
- Deployment Geography:
- Use Case:
- Release Date:
- References:
- Model Architecture:
- Input(s):
- Output(s):
- Quick Start
Headings are captured from the source. Links open the publisher’s document, not a locally hosted copy.
Links from the model card
References supplied by the publisher, not independently verified endorsements. Check the destination before downloading files or entering credentials.
- reproduction instructions
- evaluation results
- OpenMDW License Agreement, version 1.1
- MIT license and tokenizer attribution
- license
- package build evidence
- Hugging Face Transformers mBART documentation
- postprocessing examples
- the serving guide
- score sources and timing details
- downloadable results
- evaluation settings
Documentation belongs to its respective authors. Reported project/model license: openmdw-1.1. A listing is not a grant of reuse or training rights. Confirm the document’s own terms at the source.
Model overview
nvidia/Nemotron-3.5-Parse is a image-text-to-text model repository published by nvidia on Hugging Face. The source reports the transformers library.
This page summarizes Hub metadata. For intended use, training data, evaluation results and limitations, consult the original model card.
Model facts
- Task
- image-text-to-text
- Library
- transformers
- Recent downloads (30 days)
- 10,678
- Cumulative likes
- 132
- Hugging Face trending score
- 73
- Reported safetensors parameters
- 630,800,504
- Architecture
- RadioMBartPrefixForCausalLM
- License
- openmdw-1.1
- Access
- Not gated by Hugging Face
- Created
- 2026-09-03T06:29:29.000Z
- Last modified
- 2026-10-28T04:31:24.000Z
Compatibility and lineage
Language coverage was not reported.
Base-model lineage was not included.
Parameter count is not a RAM/VRAM requirement. Check precision, quantization, context length and runtime compatibility in the model card; no hardware or API-cost claim is inferred here.
Use and evaluate
Check license terms and access requirements first. Review the model card’s documented loading instructions, evaluations and safety limitations. Benchmark scores are not imported by this directory, and popularity is not an accuracy ranking.
No model weights are downloaded or executed by AltAPIs. Never enable remote model code without reviewing it.
Source and freshness
Source: Hugging Face Hub. Metadata observed 2026-10-29T06:47:43.046Z. Daily imports are snapshots, not real-time monitoring.
Popularity and source listings do not establish security, suitability, licensing rights or benchmark performance.