Search Workspace

Find roots, ayahs, lemmas, and glosses faster

Preparing corpus

You can search 2 surahs and 23 roots now; the rest of the corpus is on its way.

Quick search

Arabic roots now surface corpus-wide spread, common forms, and lemma clusters while you search.

Tokens ready32Every root, ayah, and lemma in the corpus is indexed and ready to search.

Try هدي, guidance, 2:255, or root:هدي pos:N

Current result

Select a result to keep its surah, ayah, and root context visible while you continue searching.

Corpus index

About corpus search

Search runs across all 77,429 word-level tokens, and it searches the structure rather than the letters. Ask for a root and you get every word derived from it anywhere in the text, whatever prefixes and suffixes it carries in context; ask for a lemma and you get its inflected forms.

Arabic input is normalised before matching, so a query typed without diacritics still finds the vocalised text, and the variant forms of alif, ya and ta marbuta are treated as one. Results narrow by part of speech and by morphological feature, and any hit opens in the morphology inspector with its full segment breakdown.