A guide to Sanskrit and Prakrit meters

These pages collect information (descriptions, definitions, examples) of the various meters that are used in Sanskrit and Prakrit literature.

For identifying the meter of a verse (or any chunk of text with verses in it...) I highly recommend Shreevatsa’s Metre identification tool.


Syllable-counting meters

Syllable-counting meters are characterized by a fixed number of syllables in each line.

Samavr̥ttāni (equal-pattern)
Ardhasamavr̥ttāni (half-equal pattern)


Mora-counting meters

Mora-counting meters are regulated by groups of moras and hence the number of syllables in each line is variable.



This section discusses some of the basic concepts and principles involved in Sanskrit and Prakrit meter. You might find this video (and the associated slides), which I prepared for a Prakrit seminar, to be useful.

The basic division between vr̥ttāni and jātayaḥ in the metrical repertoires of Sanskrit and Prakrit corresponds to the underlying principles of versification. The first class, vr̥ttam, refers to meters that count syllables (akṣarāṇi, varṇāḥ). The second class, jātiḥ, refers to meters that count moras (mātrāḥ).


The syllable is phonological unit. In Sanskrit it is called akṣaram or varṇaḥ. Some scholars — not me — would not like to conflate the ancient concepts associated with these words with the concept of the syllable as defined by modern linguistics. For my purposes, they are functionally identical.

A syllable consists of a single vowel (svaraḥ). It optionally has one or more consonants (vyañjanam) before this vowel, and one or more consonants after it. Therefore, using C for any consonant and V for any vowel, the structure of a syllable is C*VC* (i.e., zero or more consonants, followed by a vowel, followed by zero or more consonants).

A string of sounds is syllabified, or parsed into syllables, according to the following rules, where a period (.) represents a boundary between syllables. Note that in Sanskrit and Prakrit, syllabification generally takes place across word boundaries.

  • Each vowel goes into its own syllable.
    • ia “thus” (Prakrit) = i.a
  • A single consonant between two vowels is preferentially taken with the following vowel.
    • ēkē “some” = ē.kē
    • itō “this way” = i.tō
  • When a conjunct consonant (i.e., two or more consonants) occurs between two vowels, the first usually goes with the preceding vowel, and the last usually goes with the following vowel.
    • arthaḥ “wealth” = ar.thaḥ
    • astam “setting” = as.tam
    • jihvā “tongue” = jih.vā
  • Note: Although the principles of sonority would usually require conjunct consonants of increasing sonority (for example, tr, pr, ty) to be syllabified together with the following vowel, they are effectively ambisyllabic in Sanskrit, and the less sonorous consonant is usually doubled at least in pronunciation, if not in writing:
    • viprō “brahmin” = vip.prō
    • atyantam “too much” = at.tyan.tam
  • Note: It is useful to know that the letters called ayōgavāhāḥ in Sanskrit — namely, anusvāraḥ, visargaḥ, and their substitutes — must always come at the end of a syllable.
    • saṁsāraḥ = saṁ.sā.raḥ
    • duḥsahaḥ = duḥ.sa.haḥ

There is no conventional sign for the syllable as such in premodern metrical texts. When necessary, I represent it with the sign σ.

Moras and weight

The mora (Wikipedia) is a unit of phonological structure “below” that of the syllable, in that a syllable contains at least one mora. Intuitively, a mora represents the amount of time that it takes to pronounce something: hence a syllable with more moras is longer, or to use the terminology of both Sanskrit grammar and modern linguistics, “heavier” than a syllable with fewer moras.

In Sanskrit and Prakrit, moras are assigned to vowels and also to the consonants that follow them within a syllable, which are called coda consonants. Consonants at the beginning of a syllable do not contribute any moras. Hence, in the word prāṁśuḥ, which is syllabified as prāṁ.śuḥ, the consonants p, r, and ś have no moras associated with them. But the consononants and do. Each coda consonant is assigned one mora. Among vowels, the short vowels (a, i, u, and ) are each associated with one mora, and the long vowels (ā, ī, ū, r̥̄, l̥̄, as well as ē and ō) are each associated with two moras. Technically the long diphthongs (ai and au) are associated with three moras.

Metrical writers in South Asia did not use a symbol for the mora, but in these pages I will represent a single mora with the sign μ.

For the purposes of versification, a syllable containing a single mora is considered light, and a syllable containing two or more moras is considered heavy. The distinction between light and heavy syllables is of paramount importance for all Sanskrit and Prakrit meters, which regulate not just the number of syllables, but this quality, which is called weight. In these pages, I have represented a light syllable with the sign । (r̥juḥ, a straight line), and a heavy syllable with the sign ऽ (vakraḥ, a crooked line), in accordance with the longstanding convention in South Asia.


Prakrit meters — including those that have been borrowed into Sanskrit — are typically regulated by “groups” (gaṇāḥ). These are groups of syllables whose total number of moras adds up to a certain number, usually four (hence they are more specifically called caturmātrāḥ gaṇāḥ, “groups with four moras each”).

For parsing syllables into groups, simply count the number of moras each syllable has, stopping after the maximum of two moras per syllable has been reached. (In Prakrit there simply are no syllables that have more than two moras; in Sanskrit, however, so-called “superheavy” syllables, with three or more moras, pattern with heavy syllables.) Once you reach the required number of moras, you have a group.

Hence, where the square brackets represent the boundaries between groups, pasuvaïṇō rōsāruṇapaḍimāsaṅkantagōrimuhaandaṁ would be parsed as follows:

[][ṇō rō][sā.ru.ṇa][pa.ḍi.mā][saṅ.kan][ta.gō.ri][][daṁ]

A caesura (yatiḥ) is a place in the metrical pattern where a word-boundary is expected or required. A “word boundary” is where one word ends and another begins. The position of a caesura is usually specified with reference to syllables in the case of syllable-counting meters (e.g., “after the seventh syllables”), or with reference to groups in the case of mora-counting meters (e.g. “after the third group”). Generally word boundaries coincide with syllable boundaries (i.e., the end of a word coincides with the end of a syllable), but word boundaries that are “hidden” in the sense of being affected by sandhi are generally considered to fulfill the requirement of a caesura. Word boundaries within compound words also count for this purpose.


There were a number of overlapping traditions of metrical analysis in premodern India, and they used slightly different sets of technical terminology and conventions for describing metrical forms. Metrical writers also used the so-called bhūtasaṅkhyā system, according to which numbers were named by objects that were known to come in sets of that number.

Piṅgala’s Chandaḥsūtram

The Chandaḥsūtram is one of the earliest texts devoted to the analysis of metrical forms. It was probably compiled in the first or second century of the common era. It was known to Śabara, who perhaps wrote in the fifth century. It was also known to the author of some fragmentary texts which were found in Turfan and edited by Dieter Schlingloff. Those fragments are probably not later than about 500 CE. The Chandaḥsūtram is ascribed to Piṅgala, who was, at least by the time of Halāyudha in the ninth century, thought to be a Nāga.

Piṅgala describes both Vedic and “secular” (laukikam) meters, and also specifies a few combinatorial procedures (pratyayāḥ) for finding out information about particular categories of meters. His textbook is by far the most influential, and has been commented on a number of times:

  • the Mr̥tasañjīvanī (“what brings the dead back to life”) of Halāyudha, written in the 970s or 980s, when the poet was under the patronage of the Pāramāra king Muñja who ruled from Dhārā in present-day Madhya Pradesh;
  • the Chandōvicitibhāṣyam (“commentary on the repertoire of meters”) of Yādavaprakāśa, one of the teachers of Rāmānuja, written in the late tenth or early eleventh century in Tamil Nadu.
The triplet system

Piṅgala invented a system of “triplets” (trikāṇi), wherein one could represent a metrical pattern, consisting of a certain sequence of light syllables (।) and heavy syllables (ऽ), in an abbreviated form. In this system there are eight letters, each of which stands for a unique combination of three syllables. They can be memorized with the following mnemonic:


If we take each letter in that mnemonic and look at its own weight, in addition to the weight of the two following syllables, we arrive at the following set of equivalences:

Sign Expansion
ya ।ऽऽ
ma ऽऽऽ
ta ऽऽ।
ja ।ऽ।
bha ऽ।।
na ।।।
sa ।।ऽ

The letters la and ga stand for a single light (laghu) and a single heavy (guru) syllable, respectively.

Although these triplets were invented by Piṅgala, many subsequent authors, such as Jayadēva, Jayakīrti, Hēmacandra, Ratnākaraśānti, and Kēdārabhaṭṭa, use them in their definitions.

The Ratnamañjūṣā and Jānāśrayī

The Ratnamañjūṣā (“Chest of Jewels”) is more recent than Piṅgala’s Chandaḥsūtram, but still very old; perhaps it dates to the fourth or fifth century. It was compiled by a Jain monk who did not leave us his name. The Jānāśrayī Chandōvicitiḥ (“Janāśraya’s Repertoire of Meters”) is named for one of the Viṣṇukuṇḍin kings, who typically took the title Janāśraya; this would place it around the early seventh century. I discuss them together because they use similar conventions.

The Ratnamañjūṣā and Jānāśrayī use a slightly different system of triplets than Piṅgala. Rather than only using consonants, both vowels and consonants can stand for a particular triplet. This allows the authors to define meters in an even briefer fashion. The author of the Jānāśrayī has also included a number of abbreviations for sequences of two syllables and four syllables:

Signs (Ratnamañjūṣā) Signs (Jānāśrayī) Expansion
s ऽऽ
j ।ऽ
p ऽ।
r ।।
k, ā g, ū ऽऽऽ
c, ē , ।ऽऽ
t, au ś, ī ऽ।ऽ
p, ī l, u ।।ऽ
ś, a b, ē ऽऽ।
, u k, i ।ऽ।
s, t, ā ऽ।।
h, i m, a ।।।
d ।।।।
y ।।।ऽ
ñ ऽ।।ऽऽ

In addition to these new triplets, the authors of the Ratnamañjūṣā and the Jānāśraī also refer to the caesura in a particular way: rather than using the bhūtasaṅkhyā system, as other authors did, he names the syllable after which the caesura occurs according to the following convention:

Symbol (Ratnamañjūṣā) Symbol (Jānāśrayī) Syllable position
da na 1
di ni 3
du nu 5
dr̥ nr̥ 7
dr̥̄ nr̥̄ 8
dl̥ nl̥ 9
dl̥̄ nl̥̄ 10
dai nai 12
dau nau 14
The bhūtasaṅkhyā system

Sanskrit authors will often refer to a number by a thing that is commonly imagined to come in sets of that number. Hence “eyes” or “hands” can be used for the number 2. Many of these conventions, however, depend on a bit of cultural knowledge. One has to know that there are said to be seven “family” mountains (kulagiri-) in order to understand that expressions meaning “mountain” can express the number 7. This system is called bhūtasaṅkhyā, “entity-numerals,” i.e., using entities as numerals.

Peter Szanto has posted an extremely useful list of the terms used in this system, based on B.V. Subharayappa and K.V. Sarma, Indian Astronomy, A Sourcebook (Bombay: Nehru Centre, 1985), Appendix V. See also the Wikipedia article.

0 Words for sky: anantaḥ, antarikṣam, abhram, ambaram, ākāśaḥ, khaḥ, gaganam, jaladharapathaḥ, nabhaḥ, pūrṇaḥ, randhram, viyat, viṣṇupadam, vyōma
Others: binduḥ, śūnyam
1 Words for earth: ilā, urvarā, kuḥ, kṣitiḥ, kṣmā, gauḥ, jagatī, dharaṇi, dharā, bhūḥ, mahī, vasudhā, vasundharā
Words for moon: abjaḥ, ādiḥ, induḥ, kalādhara, kṣapākaraḥ, candraḥ, nāyakaḥ, pr̥thvī, prālēyāṁśuḥ, mr̥gāṅkaḥ, rajanīkaraḥ,vidhuḥ, śaśadharaḥ, śaśāṅkaḥ, śaśī, śītakaraḥ, śītaraśmiḥ, śītāṁśuḥ, śvētaḥ, sudhāṁśuḥ, sōmaḥ, himakaraḥ, himaguḥ, himāṁśuḥ
Others: tanuḥ “body,” rūpam “form,” pitāmahaḥ “Brahmā,” nāyakaḥ “leader”
2 Words for eye: akṣi, ambakam, īkṣaṇam, cakṣu, dr̥ṣṭiḥ, nayaḥ, nayanam, nētram, lōcanam
Words for hand or arm: karaḥ, bāhuḥ, bhujaḥ
Words for lips: ōṣṭhaḥ
Words for ears: karṇaḥ
Words for breasts: kucaḥ
Words for legs, thighs, or ankles: gulphaḥ, jaṅghā, jānu, bhujaḥ
Words for fortnights: pakṣaḥ
Words for a half-year: ayanam
Words for a pair: dvandvam, dvayam, yamaḥ, yamalaḥ, yugalam, yugmam
Words for the Aśvins: aśvī, dasraḥ, nāsatyaḥ
3 Words for fire: agniḥ, analaḥ, kālaḥ, kr̥śānuḥ, jvalanaḥ, tapanaḥ, dahanaḥ, pāvakaḥ, vaiśvānaraḥ, vahniḥ , śikhī, hutabhuk hutāśaḥ, hutāśanaḥ, hōtr̥
Words for worlds: trijagat, bhuvanam, lōkaḥ
Words for the times (past, present, and future): kālaḥ, trikālaḥ, trigataḥ
Words for the properties: guṇaḥ, triguṇah
Words for Śiva’s eyes: trinētram, haranētram
Words for cities that Śiva destroyed: puram
The word jewel (ratnam) in Buddhist writings.
The name Rāma (rāmaḥ), i.e., Balarāma, Rāmacandra, and Paraśurāma.
4 Words for ocean: abdhiḥ, ambudhiḥ, ambhōdhaḥ, ambhōdhiḥ, ambhōnidhiḥ, arṇavaḥ, udadhiḥ, jalam, jaladhiḥ, jalanidhiḥ, payōdhiḥ, payōnidhiḥ, praṇimnagēśaḥ, bandhu, lavaṇōdaḥ, varṇa, vāridhiḥ, viṣanidhiḥ, samudraḥ, salilākaraḥ, sāgaraḥ
Words for the Vedas: vēdaḥ, śrutiḥ
Words for the yugas: yugaḥ
Words for the four varṇas or āśramas: varṇaḥ, āśramaḥ
Words for the kaṣāyas among Jain authors (kaṣāyaḥ).
5 Words for arrow: iṣu, bāṇaḥ, śaraḥ, śastram, sāyakaḥ
Words for the breaths: pavanaḥ, prāṇaḥ
Words for the elements: tattvam, bhāvaḥ, bhūtam, mahābhūtam
Words for the sense faculties and their objects: akṣaḥ, arthaḥ, indriyaḥ, karaṇīyam, viṣayaḥ
The five Pāṇḍava brothers (pāṇḍavaḥ).
6 Words for systems of knowledge: aṅgam, tarkaḥ, darśanam, śāstram
Words for season: r̥tuḥ
Words for kāraka: kārakam
Words for flavor: rasaḥ
Words for the faces of Skanda: kumāravadanam, ṣaṇmukham
7 Words for horse: aśvaḥ, turagaḥ, vājiḥ, hayaḥ
Words for mountain: agaḥ, acalaḥ, adriḥ, giriḥ, nagaḥ, parvataḥ, bhūbhr̥t, śailaḥ
Words for sage: r̥ṣiḥ, muniḥ, yatiḥ
Words for day of the week: vāraḥ
Words for notes in a scale: svaraḥ
Words for Ursus Major: atriḥ
Words for the Mātr̥kā: mātr̥kā
Words for bodily fluids: dhātuḥ
Words for planet: grahaḥ
8 Words for elephant: ibhaḥ, kuñjaraḥ, gajaḥ, dantī, diggajaḥ, dviradaḥ, nāgaḥ, puṣkarī, bhūtiḥ, mātaṅgaḥ, sindhuraḥ, hastī
Words for serpent: ahiḥ, takṣaḥ, sarpaḥ
Words for the directions: dik
Words for the continents: dvīpaḥ
Words for the Vasus: vasuḥ
Words for the eight siddhis: siddhiḥ
Words for the eight maṅgalas: maṅgalam
9 aṅkaḥ, anilāhvaḥ, upēndraḥ, kēśavaḥ, gīḥ, gauḥ, grahaḥ, chidram, tārkṣyadhvajaḥ, durgā, dvāram, nandaḥ, nidhiḥ, padārthaḥ, randhram, labdham, labdhiḥ
10 avatāraḥ, aṅgulī, āśā, kakubh, karman, dik, diśā, paṅktiḥ, rāvaṇaśira
11 Words for Śiva: akṣauhiṇī, īśaḥ, īśvaraḥ, bhargaḥ, bhavaḥ, mahādēvaḥ, mr̥ḍaḥ, rudraḥ, śaṅkaraḥ, śivaḥ, śūlī, svargēśaḥ, haraḥ
12 Words for sun: arkaḥ, ādityaḥ, inaḥ, tīkṣṇāṁśuḥ, dinanāthaḥ, dinapaḥ, divākaraḥ, dyumaṇiḥ, bhānuḥ, bhāskaraḥ, maṇḍalam, mārtaṇḍaḥ, māsaḥ, raviḥ, rāśiḥ, vyayaḥ, sūryaḥ
13 aghōṣaḥ, atijagatī, karaṇam, kāmaḥ, viśvē, viśvēdēvāḥ
14 Words for Indra: indraḥ, manuḥ, śakraḥ, śarvaḥ
lōkaḥ, vidyā


This website was put together by Andrew Ollett. Most of the examples are recited by Vidvān H. V. Nagaraja Rao. I have used materials with permission from a few other sources, who are credited as appropriate:

  • Gil Ben Herut recorded H. V. Nagaraja Rao reciting a number of meters in 2005 or 2006. These recordings have been circulating in various forms for a while. For these verses, I have sometimes used the translations that Blake Wentworth drew up, and sometimes my own translations.
  • Nathan Levine recorded H. V. Nagaraja Rao reciting another set of meters in 2018. Anusha Rao uploaded these recordings to The translations of these verses are usually my own, although I have used published translations where available.

If you care about technical stuff, the data for the meters is stored in a JSON file that is read by a template. This site also uses the amazing Sanscript library for transliteration.


All of the content on this site is licensed under a CC-BY-NC-ND 4.0 license. This means that if you use it, you must attribute it to this site (or the original sources noted above), and you are prohibited from using any of this content for commercial purposes.