About FluentVerify
What this service does, what it does not do, and who made it.
What we do
FluentVerify measures how you speak English. You read a few short texts out loud. We listen, measure, and give you a report.
The report tells you two things:
- which sounds and habits make listeners work harder
- which parts of your English are already fine
What we do not do
We do not teach. We do not give lessons, drills or exercises.
Think of us like an eye test. An eye test tells you what your eyes do. It does not sell you glasses. You take the result to whoever helps you next — a teacher, a coach, or your own practice.
Why we work this way
Most people are told "your English is good" or "you have an accent". Neither sentence helps. You cannot act on it.
So we measure specific things and name them. For example: how long you pause, whether your sentence rhythm matches English rhythm, and which sounds a listener may hear as a different word.
How we measure
Three separate layers produce every number:
- Sound measurement. Software measures your speech directly — pitch, speed, pauses, and loudness.
- Pronunciation scoring. A speech service compares what you said with what the text said.
- Scoring rules. Fixed rules in our code turn those measurements into levels.
No language model writes your score. A language model only writes the sentences that explain the score, and it is blocked from mentioning any number we did not measure.
Honesty about limits
If we cannot measure something, we say so. We never guess a score.
Our levels are marked provisional. That means we have not yet finished the study that proves the exact boundary between one level and the next. We will remove that word when the study is done, and not before.