🔊 Pronunciation Reader

Type it, hear it, see exactly how it breaks apart.

Tagalog is phonemic: every letter has one sound, so syllables and pronunciation follow from the spelling with no exceptions to memorise. Type any Tagalog text below to hear it read aloud and see each word broken into syllables with a phonetic transcription. Click any word in the text to hear just that word.

Word Analysis

WordSyllablesBroad transcriptionStress type

How Tagalog Pronunciation Works

⚠️ Stress is real — and the spelling hides it

This is the one thing Tagalog spelling does not tell you. Stress is phonemic: it changes meaning. baka stressed on the first syllable (BA-ka) means cow; stressed on the second (ba-KA) it means maybe. Ordinary writing marks neither.

So this page assumes the default — stress on the second-to-last syllable — and tags every such word assumed. That default is right most of the time, but it is a guess, not a reading of the spelling.

Dictionaries resolve it with accent marks, and you can type them here:

🔤 One letter, one sound

The five vowels never shift: a /a/, e /ɛ/, i /i/, o /o/, u /u/. Compare English, where a differs in "cat" and "cake".

ng is a single letter, not two. It spells /ŋ/ — the sound at the end of English "sing" — and Tagalog uses it at the start of words too: ngayon is /ˈŋa.jon/, not "n-gayon".

r is a quick tap /ɾ/, as in Spanish pero, never the English r.

🧱 Where syllables break

Tagalog syllables are (C)V(C), and the rules have no exceptions:

Every syllable needs an onset, so a word that looks like it starts with a vowel actually starts with a glottal stop — the catch in the middle of English "uh-oh". Tagalog never writes it, but you can hear it: aso is /ˈʔa.so/, and ma·a·ga has one in the middle, /ma.ˈʔa.ɡa/.

📐 About the transcription

Unlike the French reader, this engine needs almost no exception list — Tagalog spelling is that regular. Only two words are pronounced unlike they are written: the linker ng is /naŋ/, and the plural marker mga is /maˈŋa/.

Loanword letters (c f j q v x z) are mapped to the nearest native sound, which is what most speakers do. Everything else is derived by rule.