Changelog

477,000 lemmas with 70+ fields per entry, sourced from classical dictionaries (Ibn Manẓūr, Farāhīdī) and modern corpora.

117 REST endpoints across 25 modules: dictionary, morphology, إعراب, search, Quran, TTS, ASR, embeddings, translation, and more.

TypeScript, Python, Go, and PHP SDKs — all open source and actively maintained.

Interactive API reference powered by Scalar at api.kalimalab.com/reference.

Data Points (DP) billing model — free plan includes 500 DP/day with no credit card required. Top 5,000 common words served at 0 DP.

Word of the Day endpoint — deterministic, same result across all clients until midnight UTC.

Bulk lemma lookup — up to 50 lemmas per request, capped per plan.

Full morphological analysis via CAMeL Tools: POS tagging, root extraction, conjugation tables, transliteration (Buckwalter / ALA-LC / HSB).

إعراب (syntactic parsing) — full dependency trees with classical Arabic grammar labels (فاعل، مفعول به، ...) powered by Stanza + PADT.

Quran module: 114 surahs, 6,236 ayahs, word-by-word morphology, 7 reciters, full-text search, juz/hizb/page navigation.

NLP pipeline: diacritization (Shakkala + Mishkal), dialect identification (26 Arabic cities via CAMeL MADAR), prosody meter detection.

AI layer: semantic search via pgvector HNSW, multilingual embeddings (e5-large), machine translation (NLLB-200), TTS (Azure Neural + Google Chirp3), ASR (Whisper large-v3).