PanLex is currently under development, but the database already contains more than a billion translations among 26 million expressions in about 10,000 language varieties, derived from thousands of resources. You can explore the data through several experimental interfaces, some developed by the PanLex team and some developed by others.
These interfaces are designed for human users.
This interface is designed for search engines, but anybody can use it.
To conduct your own operations on the PanLex data, you can retrieve a copy of the latest monthly archive.
If you work with Swadesh lists, look at the ones retrieved from PanLex: Swadesh 207 lists for 800 language varieties, and Swadesh-Yakhontov 110 lists for 2,000 language varieties. They are bundled into the PanLex Swadesh Corpus. It is also distributed as the NLTK PanLex Swadesh Corpora, accessible with the NLTK Word List Corpus Reader.
The code of two applications that use the PanLex API is available. You can retrieve the code for the TeraDict family (TeraDict, TümSöz, and InterVorto) and the code for PanLinx.
¹ The PanLex Tattoo Generator debuted in San Francisco at the Exploratorium Market Days mini-festival in October 02013, where dozens of visitors got temporary tattoos in (sometimes exotic) languages.
