An open, free, bilingual platform for reflection and discovery
An AI research engine continuously studies the Quranic text — its roots, arrangement and layers of meaning — and proposes hypotheses. Each one is tested by computation, measured against chance, and passed through a committee of people and AI. Every discovery that passes is published openly.
No discoveries have been published yet. The first studies are with the committee; nothing appears here without committee approval.
Test it yourself
The results below were computed by the analysis engine just now, as this page loaded.
115 = 115
The lemmas "dunya" (دُنْيا) and "akhira" (feminine singular آخِرَة) occur equally often.
Chance test: two random words of similar frequency are equal 4.2% of the time. On its own this is not proof; the counting rule and how many pairs were tried matter.
Open in Lab →19
Letters in the Basmala (simple orthography, shadda not doubled).
19 is prime — but 27% of numbers between 9 and 38 are too. A different spelling rule changes the count.
Open in Lab →×8.975
The roots of mercy (رحم) and forgiveness (غفر) occur together in 91 verses; 10.139 would be expected by chance.
Permutation test p = 0.0005 (2000 shuffles). Longer verses naturally raise co-occurrence.
Open in Lab →The research engine and visitors propose hypotheses. Each comes with verse identifiers and an explicit claim.
Numbers are counted not by the AI but by an open analysis engine. Equalities, co-occurrences and number properties are tested against chance.
Reproduces the numbers, checks grammar and root claims, and writes out alternative explanations and overreach. Its recommendation is a report, not a vote.
Members selected from Arabic, tafsir, mathematics and statistics vote independently. At least three reviews, 80% agreement and approval from a relevant expert are required.
An approved discovery is published in two languages, stating clearly what was verified and what remains interpretation, ready to share in one tap.