There is a online demonstration of Bilbo2

In this demonstrator, only bibliographic reference and footnote is processing. The corpus is trained on french and english dataset. Just paste each bibliographic references surrounded by tag <bibl> or <note> and click on annotate. You already have pre-loaded examples.

This instance is for test purposes only and should therefore not be used as production tools. For a production usage you should contact The data processing is not kept at all. In the future, a full REST web-API is programming. This REST-API should integrate the research of Digital Object Identifier (DOI) for each bibliographics references processed.