Padmini Documentation

Padmini is a Python library that generates Sanskrit words with their traditional derivations (prakriyā) in pāṇinīya-vyākaraṇa. The project aims to be:

  1. correct. If a Sanskrit word is valid, Padmini should generate it. If a Sanskrit word is not valid, Padmini must not generate it.

  2. conformant. Any prakriyā that Padmini generates should conform to the tradition. Ideally, there should be little sign that a prakriyā was generated by a computer.

  3. clean and readable. The source code should be accessible to all, and the project should be welcome to all contributors.

  4. fast. Fast code is easier to run, test, and safely modify.

As of today, Padmini can replace SanskritVerb for most use cases, but we have a few small bugs to clean up. We hope to expand Padmini to all of the Ashtadhyayi’s word types in time.


We are currently focusing on tiṅantas and subantas. Once this effort is complete, we will focus on kr̥dantas. (Last updated 25 February, 2022)


All data handled by Padmini is encoded in the SLP1 format. For details, see the SLP1 Transliteration page.

User’s Guide

API Reference