This tool converts vowelled or Biblical Hebrew into plain consonantal Hebrew by removing the nikud vowel points, with an option to also strip the te’amim cantillation accents. The result matches the unvowelled text used in most modern Hebrew writing.
How it works
Nikud and te’amim are combining marks layered on top of the consonant letters.
The tool matches them by Unicode code point and deletes them, leaving the base
letters untouched. Nikud occupy U+05B0 to U+05BC plus the shin dot
(U+05C1), sin dot (U+05C2), and qamats qatan (U+05C7). Cantillation accents
occupy U+0591 to U+05AF along with meteg, rafe, and the verse separators.
The replacement is done with two regular expressions:
nikud only -> text.replace(NIKUD_RE, "")
also marks -> result.replace(CANT_RE, "")Because only combining marks are removed, the consonant skeleton — the part that carries the word’s identity — is fully preserved.
Example
The opening of Genesis, בְּרֵאשִׁית בָּרָא אֱלֹהִים, becomes בראשית ברא אלהים
once the vowel points are stripped — exactly how the same words appear in a
modern newspaper. The counter shows how many marks were removed, which is handy
for confirming a file is fully unvowelled before importing it into a system that
expects plain text. Leave cantillation removal on for liturgical sources, since
those carry both nikud and te’amim.
Why remove nikud?
Modern everyday Hebrew — newspapers, websites, street signs, novels — is written without vowel points. Readers infer pronunciation from context, knowledge of root patterns, and word familiarity. Vowel points appear in:
- Biblical and liturgical texts — the Masoretic vocalization of the Tanakh, prayer books, and Torah study materials.
- Children’s books and language-learning materials — pointed text helps new readers who have not yet internalized enough vocabulary.
- Poetry — some poets add nikud for metrical or disambiguation purposes.
Converting pointed text to unpointed is commonly needed when:
- Importing Biblical text into a database or search system that was built for modern Hebrew, where searches against pointed text miss unpointed variants.
- Generating clean copy for a publication that mixes scholarly sources (pointed) with modern-style text (unpointed).
- Preparing text for NLP or machine-learning models trained on unpointed corpora.
- Checking that a file has been fully stripped of diacritics before submission to a system that rejects combining marks.
What the dagesh removal means
The dagesh is a dot placed inside certain letters (bet, gimel, dalet, kaf, pe, tav in Biblical grammar; also inside letters to indicate gemination). It is encoded as a separate combining character (U+05BC). Removing nikud strips the dagesh too, leaving the bare consonant glyph. In printed text the letter looks identical with or without its dagesh, but the pronunciation distinction is lost. This is expected: unpointed Hebrew text does not indicate dagesh by convention.
Difference between nikud and te’amim
Nikud (vowel points) mark pronunciation. Te’amim (cantillation accents, also called trope or trop) mark the musical phrase and stress patterns used when chanting Scripture in the synagogue. Both are combining marks that sit above or below the consonants, but they serve entirely different purposes. Some workflows need to strip nikud while preserving te’amim — for example, a cantillation study that needs the clean trope without the vowel clutter. The two toggle options address this case.