Claims Data is offered in two categories. Claims for patents granted before 2005 and after for patents granted in and after 2005. We make this distinction as a consequence of making certain improvements in parsing XML formatted USPTO provided bulk files. These improvements include improved mathematical formula parsing, retaining newlines in claim text and other improvements. Please review the updated data dictionary or reach out to with any questions.

Claims Downloads (For patents granted from 2005 and later)

202089 MB 2019456.1 MB 2018434.1 MB 2017404.7 MB 2016389.5 MB 2015374.3 MB 2014374.9 MB 2013354.5 MB
2012327.3 MB 2011301.8 MB 2010309.1 MB 2009252.6 MB 2008225.7 MB 2007238.2 MB 2006255.3 MB 2005212.6 MB

Claims Downloads (For patents granted before 2005)

2004488.6 MB 2003493.9 MB 2002482.3 MB 2001475.6 MB 2000459.8 MB 1999444.9 MB 1998433.6 MB 1997357.7 MB
1996350.3 MB 1995 330 MB 1994 325.6 MB 1993 313.1 MB 1992 308.2 MB 1991 304.7 MB 1990 291.9 MB 1989 302.6 MB
1988 258.7 MB 1987 270.4 MB 1986 226.8 MB 1985 225.8 MB 1984 206.1 MB 1983 173.6 MB 1982 175.9 MB 1981 198.7 MB
1980 185.1 MB 1979 146.5 MB 1978 195.9 MB 1977 185.9 MB 1976 194.3 MB

The PatentsView database is sourced from USPTO -provided text and XML data on published patent applications (2001-most recent update) and granted patents (1976-most recent update). The current PatentsView database MySQL dump is available for download, upon request. The patent applications database, which contains all granted and non-granted applications, is also available upon request. After March, 2016, the applications database will not contain the same inventor IDs as the PatentsView database. Only inventors on granted applications can be matched between the PatentsView and applications databases via a granted application ID.

This work was created through a government contract funded by the Office of Chief Economist in the US Patent and Trademark Office. Users are free to use, share, or adapt the material for any purpose, subject to the standards of the Creative Commons Attribution 4.0 International License (

Attribution should be given to PatentsView ( ) for use, distribution, or derivative works.

From the PatentsView database, simple assignee and lawyer disambiguations are performed, and the patents are geocoded with a location-based disambiguation. Data are then fed into the inventor disambiguation algorithm in order to identify clusters of inventor names that are determined to be the same individual. Because the disambiguation of inventor identities is an ongoing effort, there are likely to be errors observable in the PatentsView data tables. The team welcomes feedback as we continue to improve our disambiguation methodology.

For more information, visit the Methods and Sources section of the website.