QuRoots
Learn the language of the Quran — from zero to full iʿrāb.
quroots.com
QuRoots is a free platform for learning Quranic Arabic. It combines a gamified grammar course, a word-by-word Quran reader, and a searchable dictionary of every Arabic root in the Quran, so that a complete beginner can work up to parsing any verse.
Features
- 50-unit grammar course — Duolingo-style lessons with hearts, XP, streaks and checkpoint tests, taking you from the three word types (ism, fiʿl, harf) to full grammatical parsing (iʿrāb).
- Word-by-word Quran reader — all 114 surahs with translation, tafsir, and a popover on every word showing its root, form and meaning.
- Roots explorer — 1,716 Quranic Arabic roots (1,517 of them verbs) with meanings, frequencies, all ten verb forms (أبواب), conjugation tables across five tense types, and derived nouns and particles.
- Daily ayah & hadith — a small daily habit with streak tracking.
- Spaced repetition — vocabulary review scheduled by an SRS engine.
- Qirat training — recitation practice with pitch and maqam analysis.
- Leaderboards & achievements — weekly leagues, badges, and gem rewards.
Tech stack
- Next.js 14 (App Router) + React 18 + TypeScript
- Tailwind CSS 4 with CSS-variable theme tokens
- Postgres + Drizzle ORM
- NextAuth v5 (Google OAuth, database sessions)
- Framer Motion for animation, Zustand for client state
- Stripe for donations
- Deployed on Vercel
Running locally
npm install
cp .env.example .env.local # then fill in DATABASE_URL and AUTH_SECRET
npm run dev # http://localhost:3000
Production build:
npm run build
npm start
Database
Schema lives in src/db/schema.ts (Drizzle). The migrations in db/ are plain
SQL, applied in numeric order:
for f in db/0*.sql; do psql "$DATABASE_URL" -f "$f"; done
There is no migration tracking table yet — apply them in order against a fresh
database.
Content is populated by the scripts in scripts/:
npm run populate:all # surahs, ayahs, translations, words (quran.com API)
npm run populate:tafsirs
node scripts/seed-curriculum-badges.mjs
The raw morphology corpus used to build the roots, nouns and particles tables is
not distributed with this repository.
Marketing assets
Launch screenshots and social cards live in marketing/ — see
marketing/README.md for sizes and where each one goes.
Data
The Quranic linguistic data was collected from publicly available sources and
consolidated into a single structured dataset:
- 1,716 roots extracted from the Quranic corpus, 1,517 of which have verb forms
- Conjugation tables covering 10 verb forms (أبواب) and 5 tense types (ماضي، مضارع، أمر، مبني للمجهول)
- Derivational forms per bab: masdar (مصدر), active participle (اسم فاعل), passive participle (اسم مفعول)
- Quranic references — every surah:ayah occurrence for each root across all 114 surahs
Irregular Form I masdars and the derivational forms of weak roots were enriched
using the Claude API, then reviewed. The enrichment scripts are not included in
this repository.
Copyright
© 2026 QuRoots. All rights reserved.
The source code in this repository is proprietary and may not be copied,
modified, or distributed without permission. The Quranic text itself is the
revealed word of Allah and is in the public domain. The derived linguistic
dataset (conjugation tables, derivational forms, glosses, and Quranic reference
mappings) was compiled, structured, and enriched for educational purposes and is
covered by this copyright.
