How is the inherent schwa handled?

Every bare consonant in Devanagari carries an inherent /ə/. The tool adds /ə/ after each consonant unless a vowel sign (matra), a virama (्), or word-final schwa-deletion context overrides it.

What is schwa deletion?

Hindi deletes the final inherent schwa of most words, so कमल is /kəməl/ not /kəmələ/. The tool applies a word-final schwa-deletion heuristic; medial schwa deletion is more complex and left intact.

How are retroflex consonants shown?

ट ठ ड ढ ण are retroflex and surface as /ʈ ʈʰ ɖ ɖʰ ɳ/ with the retroflex hook, distinct from the dental /t̪ t̪ʰ d̪ d̪ʰ n/ of त थ द ध न.

Does it handle the nukta letters?

Yes. Nukta consonants क़ ख़ ग़ ज़ फ़ ड़ ढ़ map to /q χ ɣ z f ɽ ɽʰ/, the Perso-Arabic and flap sounds used in loanwords and many native words.

What is the Hindi IPA Transcription Tool?

Convert Devanagari Hindi to IPA in your browser, marking the full aspirated/unaspirated and dental/retroflex distinctions, nasalisation, the inherent schwa with schwa deletion, and the nukta consonants. It runs free in your browser on Gera Tools, with nothing uploaded.

Hindi IPA Transcription Tool

Name: Hindi IPA Transcription Tool
Creator: Gera Tools
License: https://creativecommons.org/licenses/by/4.0/

Hindi IPA transcription maps Devanagari to International Phonetic Alphabet symbols. Devanagari is an abugida: consonants carry an inherent vowel that vowel signs and the virama modify. Hindi’s phonology hinges on a four-way stop contrast (voiced/voiceless × aspirated/unaspirated) and a dental/retroflex split, all of which must be marked in IPA.

How it works

The transcriber walks the string, emitting a consonant’s IPA plus the inherent schwa /ə/ by default. Vowel signs replace the schwa, the virama suppresses it, and modifiers add nasalisation:

क kə   ख kʰə   ग ɡə   घ ɡʱə     (velar 4-way)
त t̪ə   थ t̪ʰə  द d̪ə   ध d̪ʰə     (dental)
ट ʈə   ठ ʈʰə   ड ɖə   ढ ɖʰə     (retroflex)
matra: ि → i, ी → iː, ु → u, े → e, ो → o …
virama ् → delete inherent schwa
ं anusvara → nasalise (ŋ/n/m by place); ँ candrabindu → vowel nasalisation ̃
final schwa deletion: कमल → kəməl

Nukta letters (क़ ज़ ड़ …) map to their Perso-Arabic and flap IPA values.

Example and notes

कमल (lotus) transcribes as /kəməl/ — two inherent schwas, with the final one deleted. ठंडा (cold) gives /ʈʰəɳɖaː/, showing the aspirated retroflex /ʈʰ/, the anusvara nasal assimilated to /ɳ/ before the retroflex /ɖ/, and the long /aː/ matra. This is a broad phonemic transcription; only word-final schwa deletion is applied, since medial deletion depends on morphology.

Hindi’s distinctive phonological contrasts

Hindi is notable among the world’s major languages for maintaining a four-way contrast in its stop consonants — a distinction that most European languages lost centuries ago and that makes Hindi phonology especially rich and challenging for non-native speakers:

Voiceless unaspirated (क /k/, त /t̪/, ट /ʈ/, प /p/) — made without voice or audible breath.
Voiceless aspirated (ख /kʰ/, थ /t̪ʰ/, ठ /ʈʰ/, फ /pʰ/) — with a strong puff of breath; confusing them with the unaspirated forms is a common learner mistake.
Voiced unaspirated (ग /ɡ/, द /d̪/, ड /ɖ/, ब /b/) — voiced without extra breath.
Voiced aspirated / breathy-voiced (घ /ɡʱ/, ध /d̪ʱ/, ढ /ɖʱ/, भ /bʱ/) — produced with a breathy or murmured quality.

Overlaid on that is the dental vs. retroflex split: Hindi has both dental stops (tongue tip touching the back of the upper teeth, marked /t̪ d̪ t̪ʰ d̪ʱ/) and retroflex stops (tongue curled back to the hard palate, marked /ʈ ɖ ʈʰ ɖʱ/). Confusing these changes word meaning.

Who uses IPA transcription for Hindi

Language learners use IPA to understand exactly how to position the tongue and how much aspiration to apply, since spelling alone does not make these contrasts clear to speakers of non-aspirating languages. Phonetics students and researchers use it to analyze data from Hindi speakers, document dialectal variation, or compare Hindi consonants to those of related Indic languages. Speech technology developers use IPA as an intermediate representation when building text-to-speech or speech-recognition systems for Hindi. Translators and subtitlers sometimes use it to produce romanized pronunciation guides for foreign-script names in Hindi texts.