Data Sources and Credits

Effective date: October 16, 2026

Lexcellent builds its word cards with the help of open linguistic data and open-source tools. We are grateful to their authors. This page lists them together with the licences that ask for credit or for derived data to be shared under the same terms.

Dictionary data — Wiktionary

Translations, pronunciations (IPA), grammatical gender, word forms and expressions come in part from Wiktionary (https://www.wiktionary.org), extracted by the Wiktextract project of Tatu Ylonen and published at kaikki.org (https://kaikki.org).

Wiktionary content is available under the Creative Commons Attribution-ShareAlike licence (CC BY-SA 4.0, https://creativecommons.org/licenses/by-sa/4.0/) and the GNU Free Documentation License. We extract, filter and reformat entries; the dictionary data derived from Wiktionary that we show is shared under the same CC BY-SA licence.

Pronunciation recordings — Lingua Libre and Wikimedia Commons

The word recordings you can play in the apps are made by native speakers who published them on Lingua Libre (https://lingualibre.org) and Wikimedia Commons (https://commons.wikimedia.org) under free licences — mostly CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/), CC BY and CC0. We play them unchanged. The speakers are credited by name on this page.

Word forms — UniMorph

Inflected word forms for Spanish, Italian and Portuguese come from the UniMorph project (https://unimorph.github.io), licensed under CC BY-SA 3.0 (https://creativecommons.org/licenses/by-sa/3.0/).

Word frequency — wordfreq

This service includes data from wordfreq (https://github.com/rspeer/wordfreq), created by Robyn Speer, used to sort words by how common they are. The wordfreq data is licensed under CC BY-SA 4.0; the software under the Apache License 2.0.

Text analysis — Stanza and Universal Dependencies

Words are split, lemmatised and tagged with Stanza (Stanford NLP Group, https://stanfordnlp.github.io/stanza, Apache License 2.0), using models trained on Universal Dependencies treebanks (https://universaldependencies.org): AnCora (Spanish), GSD (French), BOUN (Turkish), SynTagRus (Russian), Bosque (Portuguese), ArmTDP (Armenian) and the English and German UD treebanks, each under its own Creative Commons licence (CC BY or CC BY-SA).

We also use spaCy (https://spacy.io, MIT License) and simplemma (https://github.com/adbar/simplemma, MIT License).

Russian and Ukrainian word forms — pymorphy3

Russian and Ukrainian translations of adjectives are put into their dictionary form with pymorphy3 (https://github.com/no-plagiarism/pymorphy3, MIT License) and its dictionaries: Russian built from OpenCorpora (http://opencorpora.org) data, CC BY-SA; Ukrainian from the pymorphy3-dicts-uk package (https://github.com/no-plagiarism/pymorphy3-dicts), GNU GPL v3.

Translation and AI services

Translations come from DeepL (https://www.deepl.com) and Google Translate (https://translate.google.com). For some study languages (currently Italian and Ukrainian), the words of an imported text are analysed by Google's Gemini API (https://ai.google.dev). These are services we use under their terms; their output is not covered by the licences above. Machine translations and AI analysis can be wrong — please check the cards before relying on them.

Example sentences

Besides your own texts, example sentences can come from public-domain books, mostly from Project Gutenberg (https://www.gutenberg.org) and Wikisource (https://wikisource.org).

Contact

If you are an author of one of these resources and something here is incomplete or incorrect, please write to support@lexcellent.app.