What are the six tones of Vietnamese?

Vietnamese has six tones: ngang (level, no mark), sắc (rising, acute accent), huyền (falling, grave accent), hỏi (dipping, hook above), ngã (broken, tilde) and nặng (heavy, dot below). Each is shown by a distinct diacritic over or under the main vowel.

How does the tool know which tone a syllable has?

It Unicode-normalises each syllable to NFD form, separating base letters from combining marks. It then looks for the acute, grave, hook-above, tilde, or dot-below combining character and maps it to the matching tone. No mark means the level (ngang) tone.

Does it confuse vowel-shape marks with tone marks?

No. Letters like â ê ô ơ ư carry shape diacritics (circumflex, horn) that are not tone marks. The tool only counts the five tone marks plus the unmarked level tone, ignoring circumflex and horn shapes.

What counts as a syllable here?

Vietnamese is written with one tone per spelled syllable, and syllables are separated by spaces. The tool treats every whitespace-separated token containing a vowel as one syllable, stripping surrounding punctuation.

Is my text uploaded anywhere?

No. All classification happens in your browser using built-in Unicode normalisation. The text never leaves your device.

What is the Vietnamese Tone Classifier?

Scans Vietnamese text and reports how many syllables carry each of the 6 tones (ngang, sắc, huyền, hỏi, ngã, nặng), detecting the tone from the diacritic on each vowel. Runs locally in your browser. It runs free in your browser on Gera Tools, with nothing uploaded.

Vietnamese Tone Classifier

Name: Vietnamese Tone Classifier
Creator: Gera Tools
License: https://creativecommons.org/licenses/by/4.0/

Vietnamese is a tonal language in which the same string of letters can mean completely different things depending on its tone. This tool scans a passage and counts how many syllables carry each of the six tones, giving you a quick profile of the tonal texture of a text.

The six tones of Vietnamese

Each tone has a distinct name, mark, pitch contour, and phonological character:

Tone	Vietnamese name	Mark	Pitch contour	Example
Level	ngang	(none)	Mid, flat	ma (ghost)
Rising	sắc	´ acute	High, rising	má (cheek/mother)
Falling	huyền	` grave	Low, falling	mà (but)
Dipping	hỏi	̉ hook above	Mid, dipping then rising	mả (tomb)
Broken	ngã	~ tilde	High, rising with glottal break	mã (horse)
Heavy	nặng	̣ dot below	Low, falling with glottal stop	mạ (rice seedling)

The minimal set ma má mà mả mã mạ demonstrates all six tones on the same base syllable — identical consonant and vowel, six different meanings.

How it works

Every Vietnamese syllable carries exactly one tone, marked by a diacritic on its main vowel (or, for the level tone, by the absence of any mark). The classifier Unicode-normalises each syllable to decomposed (NFD) form, which splits a character such as ấ into a base a, a circumflex, and an acute combining mark. It then checks only for the five tone combining marks and maps each one:

acute  (U+0301) → sắc    (rising)
grave  (U+0300) → huyền  (falling)
hook   (U+0309) → hỏi    (dipping)
tilde  (U+0303) → ngã    (broken)
dot    (U+0323) → nặng   (heavy)
none           → ngang  (level)

Shape marks like the circumflex on â/ê/ô and the horn on ơ/ư are deliberately ignored, because they change the vowel quality rather than the tone.

What the tone distribution reveals

Analysing tone distribution across a text is useful in several contexts:

Language learning — Southern Vietnamese speech tends to merge certain tone pairs (notably hỏi and ngã are often pronounced identically in the South, while Northern Vietnamese distinguishes them). A classifier showing the ngã / hỏi ratio in a text can help learners understand regional pronunciation variation.

Proofreading and OCR quality — an unnaturally high proportion of unmarked ngang syllables in a scanned document can indicate that OCR failed to reconstruct diacritics, since tone marks are often dropped in poor-quality scans or handwriting recognition.

Linguistic research — tone frequency is not evenly distributed across Vietnamese vocabulary. Common function words and pronouns often carry low tones; verbs and nouns show more variety. A text with an unusually high proportion of high-rising tones (sắc) may reflect a specific stylistic register or high concentration of borrowed terms.

Tips for accurate results

Paste plain Vietnamese prose without embedded Latin or English text — mixed scripts can skew the ngang count since unmarked Latin tokens are classified as level-tone syllables.
Numbers and punctuation are skipped automatically.
The minimal test phrase ma má mà mả mã mạ should produce exactly 1 count in each of the six rows — use it to verify the tool before analyzing your text.