hebrew-words-db
The largest freely usable list of Hebrew words — public domain, ready for NLP and AI work.
Built with
Language
- Python
Skills & tools
- Data pipelines
- NLP
- Dataset design
- Open-source licensing
An open dataset of 67,000+ Hebrew word forms — nouns, conjugated verbs, inflected adjectives and modern slang — released under CC0 so anyone can use it, commercially or not.
The gap
Every comprehensive Hebrew word list was either AGPL — which “infects” any software that links it — or non-commercial only. That makes them unusable for most real products, open-source or commercial. Hebrew NLP deserved better.
What I built
A large, growing, plain-text list of Hebrew words: nouns in singular and plural, verbs conjugated across binyanim, tenses and persons, adjectives with gender and number agreement, adverbs, function words and current Israeli slang.
No definitions, no translations, no niqqud — just the words, because that’s exactly what was missing.
Why it matters for AI
- Tokenizer and vocabulary work for Hebrew AI language models
- Training and evaluation data for spelling, morphology and NLP tasks
- Word games and autocomplete — it powers the dictionary behind Find the Word
Released under CC0: no attribution required.