GROBID (GeneRation Of BIbliographic Data) is a machine learning library for extracting, parsing and re-structuring raw documents such as PDF into structured XML/TEI encoded documents with a particular focus on technical and scientific publications.
GROBID is an open source project supported by:
The project is maintained by Luca Foppiano, continuing the work initiated by Patrice Lopez, GROBID's original creator.
GROBID was created in 2008 by Patrice Lopez as a personal project, following a suggestion by Laurent Romary (Inria, France). In 2011, the tool was made available as open source. Since the beginning, development has been continuous, facilitated in particular by the ongoing support of Inria.
GROBID offers the following functionalities for extracting and structuring scientific documents:
GROBID is used in production by numerous major organizations and platforms:
📚 Full documentation: https://grobid.readthedocs.io
The documentation covers:
Two public demo servers are available thanks to HuggingFace:
⚠️ Warning: These demo servers are using only CRF models (faster) and have quota limitations. For serious use, please deploy your own GROBID server.
GROBID is designed for speed and high scalability:
GROBID is distributed under Apache 2.0 license.
The documentation is distributed under CC-0 license and annotated data under CC-BY license.
If you use GROBID in your research, please cite:
@misc{GROBID,
title = {GROBID},
key = {GROBID},
howpublished = {\url{https://github.com/kermitt2/grobid}},
publisher = {GitHub},
date = {2008/2026},
archivePrefix = {swh},
eprint = {1:dir:dab86b296e3c3216e2241968f0d63b68e8209d3c}
}
Contributions are welcome! If you contribute to GROBID, you agree to share your contribution under the licenses mentioned above.
For more information on contributing to the project, consult the documentation on GitHub.
GROBID is an open source project supported by INRIA and MESRE, maintained by Luca Foppiano in continuation of the work of Patrice Lopez.