Developer tools

Arabic OCR project for extracting text from document images. Benchmark claims are author-reported, not independently verified here; check each model card for weight licences.

MCP server for Arabic preprocessing in search, retrieval-augmented generation and deduplication workflows.

No-code playground and OpenAI-compatible API for humain-m3, in limited and research preview tiers.

Python toolkit for Arabizi-to-Arabic transliteration, dialect tagging and ranked candidates, with optional LLM disambiguation. README claims MIT, but licence terms remain unverified.

MCP server for Arabic dictionary lookup by word, root and meaning using Arramooz and Lane-derived data. Dataset licences are separate from the MIT server licence.

Estimates Arabic token counts and costs for GPT, Claude, Llama and Gemini. Estimates are not billing guarantees.

An open-source Arabic NLP toolkit offers morphology, entity recognition and semantic analysis.

An open-source Arabic dependency parser from CAMeL Lab (NYU Abu Dhabi). From raw text it produces tokens, part-of-speech tags, rich morphology and sentence structure in the CATiB and Universal Dependencies formats. MIT licence; needs Python 3.11.

Shakkala and Shakkelha Arabic diacritization models converted to ONNX, with a vocalize interface. Check upstream model terms too.

PyTorch ports of Shakkala and Shakkelha Arabic diacritization models. Reuse licence terms are unverified.

Elasticsearch plugin for Arabic normalization and root extraction in search.

Open-source code from SinaLab (Birzeit University) for Wojood, an Arabic named-entity model that finds 21 types of names, including people, organisations, places, events and dates, even when one name sits inside another. Trained on 550,000 words of Modern Standard Arabic and dialect text; online demo available. MIT licence.

Beta Arabic text-processing library for cleaning and parsing text, files and folders, with optional streaming.

Arabic speech toolkit for recognition, speech classification and text-to-speech, with training and pretrained models.

Research implementation of a context-sensitive neural Arabic spelling checker. Reuse licence is unverified.

Arabic preprocessing library for cleaning, normalization, segmentation and dataset splitting. Optional segmentation uses Farasa.

Arabic tokenization library with word, SentencePiece, character and other approaches.

Arabic NLP suite: morphology, named entities, sentiment and dialect ID.

Stanford NLP Group's official open-source Python library for 60+ languages. Its Arabic models split text into words and sentences, tag parts of speech, find base forms, parse sentence structure and recognise names. Apache 2.0 licence.

An open-source PHP library for Arabic websites: Arabic search queries, spell checking, sentiment and dialect detection, English–Arabic transliteration, Hijri dates, numbers in words, summarising and prayer times. LGPL v3 or Waqf Public Licence.

Research implementation of neural Arabic diacritization with models, helper tools and prediction scripts.

Neural Arabic diacritization using recurrent neural networks, with applications in speech synthesis and search.

Configurable classified Arabic stopword list with generated inflected forms in CSV, Python and SQL. Exact reuse licence terms are unverified.

Arabic dictionary for morphological analysis.

An Arabic NLP toolkit provides language-processing modules through downloadable software and REST APIs.

Snowball-based Arabic light stemmer and text normalizer, including a root-based mode. Reuse licence terms are unverified.

Arabic root-based stemmer for Apache Lucene search applications. The project licence is MIT, not Apache.

Simple embeddable Arabic spelling checker.

Software and Python library for adding diacritics to Arabic text.

Free QCRI script, first built for the IWSLT 2013 translation evaluation, that makes Arabic spelling consistent: punctuation, Arabic and Indian digits, diacritics, Alef and Ta Marbuta errors and a detached Waw. Used before scoring machine translation.

An open-source Python library that rebuilds Arabic text so letters join in their correct shapes in apps and libraries that do not support Arabic script. Pair it with python-bidi for right-to-left order. MIT licence.

Arabic morphological analyzer and lemmatizer for partially or fully vocalized text. Exact reuse licence terms are unverified.

Arabic verb-conjugation software and library with tense options and HTML, CSV, TeX and XML output.

Python library for Arabic text cleaning and normalisation.

Arabic light stemmer and segmenter with possible stem/root extraction and custom affix lists.

Lets an AI assistant manage a Zid store through a secure MCP connection.

NLP infrastructure for local and low-resource languages, including Darija.

Software technology that enables the building and deployment of software robots to emulate human interactions with digital systems.

An AI-powered data processing suite that transforms complex agricultural metrics into actionable intelligence for farmers in the Middle East.

Technological solutions focused on enabling computers to understand and interpret human language for various linguistic applications.

An integration hub that allows store owners to manage operations by connecting various AI models and agents.

A specialized spatial intelligence tool that uses AI to create high-fidelity three-dimensional digital twins of physical environments and components.

Comprehensive development environments designed for building, training, and deploying complex neural network architectures.

A measurement tool designed to track sales conversions and store visits originating specifically from ChatGPT interactions.

Open-source web app that builds Egyptian Arabic speech-recognition datasets from YouTube audio, transcribed with Gemini 2.0 Flash and exported for Hugging Face.

A developer-focused offering providing programmable interfaces for integrating predictive machine learning models into third-party applications.

AI systems designed to derive meaningful information from digital images and videos for identification and analysis.