OUR MISSION

To make the Buddha’s teachings, as preserved in the Pāli Canon, accessible to people around the world.

OUR VISION

A world in which each person can study the Buddha’s teachings in their own language and apply them on the path to Nibbāna.

THE CHALLENGE

The Pāli Canon (Tipiṭaka) and its commentarial literature contain approximately 10 million words. English translations of the main texts have been produced over the last 150 years by different translators, in different periods, with different terminology. As a result, the same Pāli term is often rendered in several ways, the translations are spread across many sources, and large parts of the commentarial literature have not been translated at all.

PaliVerse was created to bring the Canon and its commentaries together in one place, in a single translation that renders each Pāli term consistently throughout.

HISTORY

• 2018. Michael Xynos founded the Centre for the Study and Practice of Theravāda Buddhism in Athens, a non-profit organisation. One of the aims in its articles of incorporation was to translate the Pāli Canon into English, Greek and other languages. The means to carry out this aim did not exist at that time.
• 2019. At the invitation of Michael Xynos and Venerable Ratana Mahāthera, head of the Sri Lankan Theravāda community in Greece, Most Venerable Ñāṇadassana Mahāthera came to Athens. He is a Greek monk of the forest tradition who has studied the Tipiṭaka for several decades. He taught at the Centre for six years, giving Dhamma talks and meditation courses to the Greek public. His teaching was based directly on the Pāli texts. He is the author of about twenty books, and the Centre published several of them in Greek. His text-based method became the foundation of PaliVerse.
• 2022. Large Language Models became publicly available, and the vision set out in 2018 appeared within reach. In their first years, however, these models translated poorly, often worse than Google Translate. About two years later, they had improved enough to translate texts into English even from low-resource languages such as Pāli.
• October 2024. The first experimental translations began. PaliVerse applied standard technology that the translation and localisation industry has used for decades: computer-assisted translation tools, text segmentation and translation memory. To this was added a glossary of approximately 6,000 Pāli terms, prepared with Venerable Ñāṇadassana, so that the terminology would be fixed before full-scale translation. The translation memory ensures that each term, and each passage that repeats across the texts, is translated the same way everywhere. With this method, a small team could undertake work that would otherwise have required many translators over several generations.
• 2025. In early 2025, the first version of the English translation of the Pāli Canon was completed. Since then, it has been continuously revised and improved. In the same year, Venerable Ñāṇadassana returned to Sri Lanka. He continues to guide PaliVerse and oversees the revision of the English translation.
• 2026. The PaliVerse team developed the PaliVerse AI Corrector, a tool designed to work together with human experts to accelerate the correction of the translations. It follows the well-established human-in-the-loop approach to AI systems, in which human experts guide and oversee the work.

CURRENT STATUS OF THE TRANSLATIONS

English

• Root text (Mūla) of the three Piṭakas (Sutta, Vinaya, Abhidhamma): translated.
• Commentaries (Aṭṭhakathā): translated.
• Sub-commentaries (Ṭīkā): translation is about to begin.
• The root text translation is under extensive revision, overseen by Venerable Ñāṇadassana. The revision includes work on the consistency of the terminology.

Greek

• Root text: translated.
• Commentaries and sub-commentaries: not yet translated.

Other languages

• Translation into more languages is planned.

HOW THE TRANSLATIONS ARE MADE

The translations are produced by a Large Language Model (Anthropic’s Claude) within a professional translation workflow:

• A glossary of approximately 6,000 Pāli terms, developed with Venerable Ñāṇadassana, was prepared before translation and is applied across all texts.
• Text segmentation and translation memory are managed in Trados Studio, a computer-assisted translation tool.
• Each text is translated directly from Pāli into the target language, without an intermediate language.
• The output passes through automated correction tools and review by human experts.

The translations are machine-generated and remain under revision. Errors are possible, and corrections from scholars and readers are welcome.

READING THE TEXTS

The root text and the commentaries are linked at paragraph level, so a reader can move from a numbered paragraph of the root text to the commentary on it.

ASK PALIVERSE

Ask PaliVerse is a question-and-answer tool. It retrieves passages from the translated texts before it generates an answer, and it gives references to those passages.

The version now online is the first version. It covers the suttas and their commentaries only. It does not yet cover the Vinaya, the Abhidhamma, or the sub-commentaries, so its references on Vinaya and Abhidhamma questions may be inaccurate. A second version covering the whole Tipiṭaka is in preparation.

ACCESS AND FUNDING

PaliVerse is operated by a non-profit organisation. All content, including Ask PaliVerse, is free to use. At present, the project is funded mainly by its founder. We hope that, in the future, donations will support its further development and its long-term sustainability.

PLANNED WORK

• English translation of the sub-commentaries (Ṭīkā).
• Completion of the revision of the English root text.
• Translation into more languages.
• The second version of Ask PaliVerse.

CONTRIBUTING

• Scholars and Pāli readers can send corrections and comments on the translations.
• Donations support the running of the platform and further translation work.
• Institutions interested in cooperation can contact us by email at [email protected].