Hebrew Nlp Toolkit
Verified97/100Guide developers in using Hebrew NLP models and tools including DictaLM, DictaBERT, AlephBERT, and ivrit.ai. Use when user asks about Hebrew text processing, Hebrew NLP, "ivrit", Hebrew tokenization, Hebrew NER, Hebrew sentiment analysis, Hebrew speech-to-text, or needs to process Hebrew language text programmatically. Covers model selection, preprocessing, and Hebrew-specific NLP challenges. Do NOT use for Arabic NLP (different tools) or general English NLP tasks.
Trust score 97/100 (Verified) · 1,934+ installs · 2 GitHub contributors · MIT license
Natural language processing for Hebrew remains a significant technical challenge. Hebrew is a morphologically rich language with unvocalized script, making tasks like entity extraction, sentiment analysis, and part-of-speech tagging far more complex than in English. Available models are scattered and not always well documented.
skills-ilHow to use this skill
Not sure how? Read the guide- 1. Click "Download ZIP" to download the skill files.
- 2. Open Claude Desktop and go to Customize > Skills.
- 3. Click "+" and select "Upload a skill", then upload the ZIP file.
- 4. Start a new conversation. The skill will activate automatically when relevant.
Developers? Install via command line (CLI)
npx skills-il add skills-il/localization@v1.3.0-hebrew-nlp-toolkit --skill hebrew-nlp-toolkit -a claude-codeWhen to Apply
- When tokenizing and morphologically processing Hebrew text
- When performing Named Entity Recognition (NER) on Hebrew to extract names, places, organizations
- When using Hebrew BERT models like DictaBERT or AlephBERT for embeddings
- When restoring nikud (diacritization) to unvocalized Hebrew text with Dicta Nakdan
- When handling Hebrew morphology challenges (binyanim, inflections, smichut)
Try These Prompts
How do I use DictaBERT for sentiment analysis of Hebrew texts? Provide a complete Python code example.
How do I perform named entity recognition (NER) in Hebrew using DictaBERT NER? I want to identify people, places, and organizations in text.
How do I restore nikud (diacritization) to unvocalized Hebrew text? Show me both the Dicta Nakdan model and the hosted API option.
What are the best ivrit.ai tools for Hebrew speech-to-text conversion? How do I integrate them into a Python project?
Frequently Asked Questions
Changelog
Fixed the DictaBERT load snippet (base masked-LM via AutoModelForMaskedLM, not a randomly-initialised SequenceClassification head; points to dictabert-sentiment/ner for classification) and reframed the ivrit.ai 22K-hour corpus claim.
Jun 16, 2026
Added nikud restoration, hosted Dicta APIs and alternative frameworks; corrected the DictaLM 3.0 base-model lineage and tech-report link.
May 14, 2026
Corrected DictaLM 3.0 sizes (24B, 12B Nemotron, 1.7B), verified HuggingFace model IDs, fixed VRAM requirements, added Reference Links section, and included NeoDictaBERT bilingual embeddings.
Apr 14, 2026
Related Skills
Author: skills-ilGenerate professional Hebrew documents (PDF, DOCX/Word, and PPTX) with correct right-to-left layout, mixed Hebrew-and-English bidi handling, and proper Hebrew typography. Use whenever the output is a Hebrew or mixed Hebrew/English Word document, Hebrew PDF, or Hebrew PowerPoint, including phrasings like "Hebrew Word document", "Word document in Hebrew", "מסמך Word בעברית", "create a .docx in Hebrew", "lehafik heshbonit", and "litstor hozeh", or Israeli templates such as Heshbonit Mas (tax invoice), Hozeh (contract), Hatza'at Mechir (proposal), or Protokol (meeting minutes). ALSO use this for the symptom where a Hebrew document looks correct on screen or in Claude but comes out scrambled, reversed, or broken after export to Word, with English words, numbers, or punctuation landing on the wrong side, phrased as "Hebrew text reversed in Word", "my Hebrew Word file is broken", "fix Hebrew formatting in Word", or "the docx came out messed up"; the fix is regenerating the .docx with paragraph-level RTL/bidi, NOT a web/CSS RTL change. Prefer this over the generic docx or pdf skills ONLY when the document is Hebrew or right-to-left, because those do not set RTL/bidi and produce scrambled Hebrew with English words and punctuation in the wrong place; for English-only documents use the generic skill. Covers reportlab, WeasyPrint, python-docx, and pptxgenjs. Do NOT use for OCR or reading existing documents (use hebrew-ocr-forms instead).
Author: skills-ilWrite and edit professional content in Hebrew including marketing copy, UX text, articles, emails, and social media posts. Use when user asks to write in Hebrew, "ktov b'ivrit", create Hebrew marketing content, edit Hebrew text, write Hebrew UX copy, or optimize Hebrew content for SEO. Covers grammar rules, register from formal to dugri, mixed Hebrew/English, gendered language, nikud and numerals, and Hebrew SEO best practices. Do NOT use for Hebrew NLP/ML tasks (use hebrew-nlp-toolkit) or translation (use a translation skill).
Author: skills-ilImplement right-to-left (RTL) layouts for Hebrew web applications. Use when user asks about RTL layout, Hebrew text direction, bidirectional (bidi) text, Hebrew CSS, "right to left", or needs to build a Hebrew web UI. Covers CSS logical properties, the :dir() pseudo-class, Tailwind RTL, React/Next.js RTL setup, icon mirroring, Hebrew typography, and font selection. Do NOT use for Arabic RTL (similar but different typography) unless user explicitly asks for shared RTL patterns, or for native mobile RTL (React Native I18nManager, SwiftUI, Android) which is out of scope.
Use at your own risk. Terms of Use · Security
Want to build your own skill? Try the Skill Creator · Submit a Skill