ChabadLabs AI FOR SHLICHUS

Sefaria

The largest digitized Torah library; the source-of-truth corpus AI should be reading from.

On this page

The largest digitized Torah library. Not AI itself, but the corpus that AI most often should be reading.

What it is

sefaria.org is an open-source library of Torah, Talmud, Midrash, halacha, and Chassidus, with JPS 1917, Metzudah, and other translations. Kehot put Tanya and Likutei Torah on Sefaria as of February 2026.

How the chat uses it

  • Source-of-truth for citations. When ChatGPT cites a Gemara page, paste the citation into Sefaria's exact-match toggle to verify before quoting.
  • Translations. Multiple translation versions per text; pick the one fitting your audience.
  • Paste primary text into AI. Rather than asking AI to fetch a Mishnah from memory, copy the Mishnah from Sefaria into your prompt.
  • Translation choice for the Megillah App. JPS 1917 felt archaic, so the builder switched to Metzudah, with Kehos pending approval.

Why it matters

"Tried ChatGPT, Claude, and Google to find a specific Gemara reference. ChatGPT and Google both wrong. Claude was honest about limitations and recommended Sefaria (which worked)."

This is the chat's repeated lesson: Sefaria for the text, AI for the reading. Do not reverse it.

  • Dach.dev, a plain-text Chassidic library with copy-full-sicha.
  • HebrewBooks.org, a large repository of Hebrew sources, with less AI access.
  • anash.org Likkutei Sichos app, plain-text Likkutei Sichos.
  • likuteysichos.com, topic search.
  • rav.dicta.org.il, Hebrew NLP for Torah.
  • chabad.org, translations and articles.

Last updated May 2026 · Maintained by Hermes AI Agent