Short answers to the questions readers, researchers, and developers ask most. If something isn't here, email us.
Neither, and both. The corpus includes classical primary sources from both Sunni and Shi'a traditions because serious research into Islamic history requires access to both. When the two traditions agree, we show the agreement. When they disagree, we return both positions with each side's reasoning. We don't flatten disagreements, and we don't pick a side for you.
An independent developer, not a scholar. This is infrastructure work — the same category as a library catalog or a search index — built by someone passionate about making the classical Islamic canon accessible. Religious authority belongs to qualified scholars, not to software.
No. Islamic Corpus is a research tool. It retrieves passages from classical sources and summarizes what those sources say. It does not issue rulings, give religious advice, or speak on behalf of Islam. For fiqh questions, personal religious matters, and the application of Islamic teaching to your life, consult a qualified living scholar.
Because the system is retrieval-grounded. Every answer is built from passages that already exist in the corpus — we don't generate claims out of the model's training data. Each response shows you the source it's drawing from: the named scholar, the volume, and the passage text. If Tabari didn't write it, we can't claim he did, because the passage itself has to come from somewhere in the indexed text.
Over 200 classical primary sources spanning sira, hadith collections, classical histories, tafsir, geography, and regional chronicles from al-Andalus to Southeast Asia. The full list is on the sources page, with each source tagged by era, tradition, and chain strength. New sources are added regularly.
The corpus focuses on the classical scholarly tradition surrounding the Qur'an — tafsir, hadith, sira, and history — rather than the Qur'anic text itself. Excellent tools already exist for searching the Qur'an directly (Quran.com, Tanzil.net, and others). We index what the scholars wrote about it.
Not currently. The corpus is the classical canon — broadly, works from the early Islamic period through roughly the 16th century, with some extension into later classical commentaries. Modern scholarship (20th–21st century) is a different project and is outside our scope.
No. Ask questions in plain English and receive answers in English. Original Arabic terms and names are preserved where they matter — technical vocabulary, proper nouns, and titles of works — so you can trace terminology back to the source if needed.
Cite the primary source, not the tool. Islamic Corpus surfaces the citation — the named scholar, the work, and the volume — but the scholarly authority belongs to the original author. We're the index; Ibn Kathir is the source.
No. Queries and API usage are not used to train any AI models. The corpus itself is the knowledge base; your questions are processed to retrieve from it and are not retained for model improvement.
Confirmation emails are sent via Resend using noreply@islamiccorpus.com. If you don't see one within a couple of minutes, check your spam or junk folder first. Outlook and Hotmail in particular filter transactional emails more aggressively than other providers, and sometimes deliver confirmation emails to spam or drop them silently. If the email still doesn't appear, signing up with a Gmail address usually works. You can also contact us if you're blocked — we can manually confirm the account.
Your query text is not stored in a way that is tied to your identity or shared outside the system. We do retain anonymized signals about how the corpus performs — which passages were retrieved, how often queries could not be answered, which figures and eras are well or poorly covered — and we keep those signals indefinitely. They're what we use to improve retrieval quality and identify gaps in the corpus. No personal content, no conversation history linked to you, no data sold to third parties.
Each source in the corpus is split into passages and indexed as semantic embeddings. When you ask a question, the system converts your query into the same embedding space, retrieves the most relevant passages across all indexed sources, and passes them to a large language model for synthesis into a readable answer. The model does not generate claims from its training data — it summarizes from the retrieved passages only. Every answer includes the named scholar, the work, and the volume the passage came from, so you can verify directly.
Each tier includes a monthly quota of API and chat queries, shown on the pricing page. Within a tier, short bursts of requests are allowed; sustained usage beyond the monthly quota returns a clear error rather than silently degrading. Institutional tier queries are unmetered on the API side. If you need higher limits or a custom arrangement, get in touch.
Commercial use is covered by the Institutional tier. That includes apps, products, and services that are charging users, resellers, or any deployment where Islamic Corpus is part of a paid offering. Individual research, academic work, personal projects, and non-commercial community tools are covered by the Free, Researcher, and Developer tiers. If you're unsure which category you fall into, email us and we'll figure it out.
New sources are added regularly as ingestion and authentication review complete. Active expansion areas include regional chronicles from al-Andalus, Ottoman and Mughal history, West and East African Islamic scholarship, and Southeast Asian traditions. The homepage stats reflect the current state of the corpus and update as the database grows.
Every tier includes the full corpus. The free tier is designed to actually be used — not to frustrate you into paying.