About Quran.DIY

Quran.DIY helps you study the Quran through the Quran itself. Ask a question in plain language and it finds the verses that speak to it, groups them by theme, and shows how the text explains its own concepts.

Why only the Quran?

This is a research choice, not a claim about how the Quran should be understood. By drawing every answer from the Quranic text alone — with no tafsir, hadith, or later commentary mixed in — the tool lets you see plainly what the text itself says on a topic before any interpretation is layered on top. That kind of clarity is useful to everyone: students, researchers, and anyone tracing how a theme runs across the whole book. The Quran is the latest revelation, and I believe it deserves deep, first-hand study from each of us — that conviction is why this tool exists.

How This Application Works

This application uses AI technology to help you discover relevant Quranic verses and thematic connections. All responses are generated based solely on the Quran text, our curated topic database, and knowledge graph of Quranic entities. No external religious interpretations, commentaries, or non-Quranic sources are included in the AI responses. The AI chat feature searches exclusively through Arabic text to ensure accuracy and authenticity. If you want to search for specific words or verses in translations, please use the Chapters feature instead.

Search & Discovery Features

Beyond asking questions, you can interact with Arabic text by selecting individual words or phrases from search results and adding them to your search. The app includes a comprehensive hierarchical topic system with 2,200+ topics and 390+ subthemes, plus a beta knowledge graph connecting people, places, and concepts mentioned in the Quran.

Beta Knowledge Graph

Our knowledge graph feature extracts and connects entities, relationships, and concepts from all Quranic verses using AI analysis. This experimental feature helps visualize connections between people, communities, and divine relationships mentioned in the Quran. Please note that AI analysis may contain ambiguities or errors. We welcome your feedback and suggestions through our Contact form or directly at [email protected].

Technology

Answers come from a semantic search engine (PostgreSQL with the pgvector extension) over the Arabic text and a curated topic set, together with a knowledge graph (Neo4j) of the people, groups, and relationships the Quran names. Meaning-matching uses OpenAI embeddings. The topic and entity layers were extracted independently by two models — GPT-5.6-Terra and GPT-5.6-Luna — run as an A/B pair, keeping only what both models agree on, so no single model's reading stands on its own. Built with Laravel.

Developed with the help of Claude Code (Opus 4.8) as a pair-programming tool.