Q: What makes the Document Understanding Transformer different from traditional OCR tools?

A: The Document Understanding Transformer is "OCR-free," meaning it does not require a separate optical character recognition engine to preprocess text. Instead, it uses an end-to-end Transformer model to recognize and understand text directly from images.

Q: Is the Document Understanding Transformer free to use?

A: Yes, the software is released under the MIT License, allowing for broad usage and modification.

Q: Where can the technical paper for Document Understanding Transformer be found?

A: The model is described in the paper "OCR-free Document Understanding Transformer," which is accessible via arXiv (ID: 2111.15664).

Document Understanding Transformer

OCR-free end-to-end Transformer model

Intangible transformer Q128801599

Facts

10 properties

Instance of: transformer
Official website: arxiv.org^[10]
License: MIT License
Website: arxiv.org
Has use: optical character recognition^[9]
Copyright license: MIT License^[2]
Described by source: OCR-free Document Understanding Transformer^[11]
Software version identifier: 1.0.3, 1.0.4, 1.0.5, 1.0.7, 1.0.8, 1.0.9^[3]^[4]^[5]^[6]^[7]^[8]
Use: optical character recognition
Version: 1.0.3, 1.0.4, 1.0.5, 1.0.7, 1.0.8, 1.0.9

🏷️ Also known as

en Document Understanding Transformer en donut

🔖 Identifiers · 2

Described at URL: huggingface.co
Source code repository URL: github.com

Report an error or request removal Lineage (PROV-JSON)

🔗 Connections

MIT License

Has use 1

optical character recognition

Instance of 1

transformer

← Based on 1

Nougat

Press Enter · cited answer in seconds

Summary

Document Understanding Transformer is a transformer^[1].

References

Programmatic citations — every numbered marker resolves to a verifiable graph row below.

Direct Wikidata claims

[2] ↑ Document Understanding Transformer — copyright license (P275): MIT License. api.github.com. Retrieved 2024-08-11. api.github.com. Provenance: wikidata.org.
[3] ↑ Document Understanding Transformer — software version identifier (P348): 1.0.3. github.com. Retrieved 2024-08-11. github.com. Provenance: wikidata.org.
[4] ↑ Document Understanding Transformer — software version identifier (P348): 1.0.4. github.com. Retrieved 2024-08-11. github.com. Provenance: wikidata.org.
[5] ↑ Document Understanding Transformer — software version identifier (P348): 1.0.5. github.com. Retrieved 2024-08-11. github.com. Provenance: wikidata.org.
[6] ↑ Document Understanding Transformer — software version identifier (P348): 1.0.7. github.com. Retrieved 2024-08-11. github.com. Provenance: wikidata.org.
[7] ↑ Document Understanding Transformer — software version identifier (P348): 1.0.8. github.com. Retrieved 2024-08-11. github.com. Provenance: wikidata.org.
[8] ↑ Document Understanding Transformer — software version identifier (P348): 1.0.9. github.com. Retrieved 2024-08-11. github.com. Provenance: wikidata.org.
[9] ↑ Document Understanding Transformer — has use (P366): optical character recognition. wikidata.org.
[10] ↑ Document Understanding Transformer — official website (P856): https://arxiv.org/abs/2111.15664. api.github.com. Retrieved 2024-08-11. api.github.com. Provenance: wikidata.org.
[11] ↑ Document Understanding Transformer — described by source (P1343): OCR-free Document Understanding Transformer. wikidata.org.

Class ancestry

[1] ↑ Document Understanding Transformer is an instance of transformer (Q85810444) [P31, primary]. Wikidata. wikidata.org.

📑 Cite this page

Use these citations when quoting this entity in research, articles, AI prompts, or wherever provenance matters. We aggregate Wikidata + Wikipedia + authoritative open-data sources; the stitched, scored, cross-referenced view is what 4ort.xyz contributes.

APA

4ort.xyz Knowledge Graph. (2026). Document Understanding Transformer. Retrieved October 25, 2026, from https://4ort.xyz/entity/document-understanding-transformer

MLA

“Document Understanding Transformer.” 4ort.xyz Knowledge Graph, 4ort.xyz, 25 Oct. 2026, https://4ort.xyz/entity/document-understanding-transformer.

BibTeX

@misc{4ortxyz_document-understanding-transformer_2026, author = {{4ort.xyz Knowledge Graph}}, title = {{Document Understanding Transformer}}, year = {2026}, url = {https://4ort.xyz/entity/document-understanding-transformer}, note = {Accessed: 2026-10-25}}

LLM prompt

According to 4ort.xyz Knowledge Graph (aggregator of Wikidata, Wikipedia, and authoritative open-data sources): Document Understanding Transformer — https://4ort.xyz/entity/document-understanding-transformer (retrieved 2026-10-25)

Canonical URL: https://4ort.xyz/entity/document-understanding-transformer · Last refreshed: October 25, 2026