← All guides

Why your pronunciation is not improving — phoneme-level scoring explained

5 min read · Updated 2026-08-06

You can practise pronunciation for months and hear no difference. Usually the problem is not effort — it is that nothing tells you what is wrong. A score of 78 does not say which sound cost you the points. Only when sounds are judged one by one does the fix become visible. Tokkly scores every phoneme separately when you read a sentence aloud, and names the weak ones along with tongue and lip position guidance for fixing them. Japanese sound length, Chinese tone (shown only as a hint, not scored), and Korean final consonants each get separate handling, and pronunciation scoring is unlimited on the free plan.

Why one score for a sentence does not help

Most pronunciation tools compare your whole recording against a native reference and return a single number. When it is low, the only possible advice is "try again". The part that matters — which sound went wrong — is missing.

So the same mistake repeats. A Korean speaker mixing English R and L, or turning F into P, is not being careless: those sounds do not exist in the first language, so the habit is systematic. Systematic problems need to be named to be fixed.

What phoneme-level scoring does

A phoneme is the smallest unit of sound that changes meaning. Only one sound separates "rice" from "lice".

Phoneme-level scoring lines your speech up against the sounds the sentence requires, and scores each one separately. Speech recognition asks "what did you say"; this asks "did you make that sound". They are different questions — speech can be perfectly understandable and still miss a phoneme every time.

That is how Tokkly works. Read a line aloud and you get a score per sound, with the weak ones collected for you. What to fix arrives with a name attached, like "your R is weak".

The real screen
“appreciate”
ə88p90r84ʃ44eɪ86

A score is not enough — you need the how

Knowing which sound is weak does not help if you do not know how to make it. So each sound comes with where to put your tongue and lips, and which sound in your own language it gets confused with.

English R, for example, is not Korean ㄹ. In ㄹ the tongue tip touches the roof of the mouth; in English R it touches nothing. Knowing that one line is the difference between months of progress and none.

The real screen

What changes per language

In Japanese, length changes meaning (おじさん uncle vs おじいさん grandfather), so vowel and consonant length is judged separately.

In Chinese, tone changes meaning — but tone mixes with sentence intonation and is hard to measure reliably in short utterances. Tokkly therefore shows tone as a hint rather than a score: withholding judgement beats confidently telling you something wrong.

For learners of Korean there is a separate model tuned to Korean sounds, because general models handle positions like the final consonant (받침) poorly.

A practice loop that works

1. Read a short sentence aloud and get per-sound scores.

2. Pick only one or two weak sounds. Trying to fix everything fixes nothing.

3. Drill words containing just that sound, many times.

4. Go back to the sentence. Sounds get fixed inside words and settle inside sentences.

Common questions

Do I have to pay to use pronunciation scoring?

No. Pronunciation scoring is unlimited on the free plan.

Do I need to sound like a native speaker to score well?

No. The goal is not imitation — it is making sounds distinct enough that meaning does not change. A noticeable accent with clear phonemes communicates well.

Do I need a special microphone?

The built-in microphone on a phone or laptop is fine. A quiet room matters far more than the hardware.

Score your pronunciation free →No sign-up needed

Read next

You understand it but cannot say it — the production gapReading and listening improve without speaking following along. Here is why comprehension and production are separate skills, and what to change in your practice.How to practise speaking a language alone, for freeYou do not need a class or a native-speaker friend to improve speaking. Here is what is missing when you practise alone, and a routine that fills those gaps.You memorised it but cannot say it — storing versus retrievingWhy the vocabulary you drilled never surfaces in conversation, and the retrieval practice that turns memorised words into usable ones.Talking with an AI tutor — what actually happens when you hold the micSpeaking practice usually stalls because there is nobody to talk to. Tokkly's AI tutor understands you when you speak, and scores your pronunciation afterward.⚡Speed Speak — why speaking against a timer is what makes it stickA memorised word and one that comes out on cue differ. Speed Speak shows only the meaning and makes you say it aloud before time runs out — exposing that gap right there.How Tokkly's adaptive skill test decides how hard each question isThe same questions for everyone miss a strong learner's ceiling and a weak learner's real level. Tokkly's test is adaptive (CAT), so five minutes reveals your level per skill.How to do shadowing — and the one thing it will never tell youShadowing builds rhythm and speed fast, but nothing tells you what you got wrong while you are doing it. Here is the routine, and how to fill that gap.Talking to yourself in a foreign language — why it works, and where it quietly stopsTalking to yourself is the cheapest speaking practice there is — but nobody corrects you, and you reuse only words you already know. Here is the fix.TOEIC Speaking on your own — what studying alone actually leaves outTOEIC Speaking is spoken and timed. Studying alone leaves two gaps: nobody tells you whether you were understood, and you never rehearse the clock.