Pali Canon Search

About this search

A quiet tool for people who study the early Buddhist texts: find a passage by the words it uses or by what it is about, read it with the Pāli alongside, and see where else in the canon the same teaching appears.

What is here

4,167 texts from the Sutta Piṭaka of the Pāli Canon (the four main nikāyas and the books of the Minor Collection that Bhikkhu Sujato has translated), 135,499 segments in all. Each text is shown segment by segment, so any English sentence can be lined up with its Pāli.

The Vinaya, the Abhidhamma, and the Minor Collection books without a Sujato translation are not included yet.

Where the texts come from

English translation
Bhikkhu Sujato, dedicated by him to the public domain under Creative Commons Zero (CC0).
Pāli text
The Mahāsaṅgīti edition, as published by SuttaCentral; see their licensing page for its terms.
Source
SuttaCentral, from their bilara-data repository (commit 738f45ee77, 2026-07-31). Every text page links back to the same text on SuttaCentral. Errors in layout or search are ours, not theirs.

With deep thanks to Bhikkhu Sujato and everyone at SuttaCentral, whose decision to give these translations away is what makes a tool like this possible.

How the search works

Words
A full-text index over the English, matching word forms (craving finds craves). Built with Pagefind. Results are passages of about a paragraph, ranked by how well they match.
Meaning
Every passage (9,117 of them) was turned ahead of time into a list of numbers that captures what it is about, using the open bge-small-en-v1.5 language model. When you search by meaning, your browser downloads that same small model once (about 55 MB with its runtime), turns your question into numbers the same way, and finds the passages closest to it. It finds passages about what you asked even when they use other words. It can also be wrong in ways a word search is not, so read what it finds.
Pāli
A full-text index over the Pāli root text.
Similar passages elsewhere
At the foot of each text, the passages in other texts whose meaning sits closest to a passage in this one (24,947 links across the canon). These are found by the model, not taken from the traditional lists of parallels, so they include true parallels, shared stock passages, and the occasional coincidence.

Privacy

There is no server behind this site, only files. Your searches run in your browser. What you type stays in the part of the address after the #, which browsers never send to a website. Meaning search runs the model on your own device and fetches nothing that depends on your question. Word search fetches the small pieces of the index it needs, and the host serving these files can see which pieces were fetched, which hints at the words searched. There are no accounts, no cookies, no analytics and no third-party requests of any kind.

Reuse

The translations are in the public domain, so you may copy, quote and reuse them freely. Credit to Bhikkhu Sujato and SuttaCentral is the kind thing to do.

Who made this

Built by Jim Irving with Claude, an AI model made by Anthropic, as an offering to anyone who studies these texts. All the translation and editorial work is Bhikkhu Sujato's and SuttaCentral's; this site only makes it searchable.