How Prepyo scores your writing and speaking
Marking criteria are public documents. Here is which ones your answers are measured against, what the model looks at, and where an estimated score stops being useful.
The criteria your answers are marked against
Pearson and the IELTS partners publish the descriptors their examiners work from. Prepyo scores against those same criteria rather than a scale of its own. Each card links to the official document, so you can check them yourself.
IELTS Academic — Writing
- Task Achievement / Task Response
- Coherence and Cohesion
- Lexical Resource
- Grammatical Range and Accuracy
The four are equally weighted and averaged, so the weakest one caps the band.
Official source
IELTS Academic — Speaking
- Fluency and Coherence
- Lexical Resource
- Grammatical Range and Accuracy
- Pronunciation
Assessed across all three parts of the speaking test, not part by part.
Official source
PTE Academic
- Communicative skills: listening, reading, speaking, writing
- Content, form and task fulfilment
- Oral fluency and pronunciation
- Grammar, vocabulary, spelling and written discourse
Most tasks are integrated, so one answer contributes to more than one score.
Official source
What the model actually measures
Speaking
Your recording is transcribed and compared with the expected response. Each word gets a pronunciation score, and your pace, hesitations and pause placement are measured — the three things a fluency mark is built from and the three you cannot judge in your own voice.
Writing
Your text is assessed against each criterion separately rather than as one impression, so the report can tell you that structure is holding you back while vocabulary is not. Individual sentences come back with a stronger version and the reason it is stronger.
Reading and listening
These are marked objectively against the answer key, the same way the real test does it. There is no estimation involved — a right answer is a right answer.
What an estimated score cannot tell you
This section matters more than the one above it. A prep score that is treated as a promise does more damage than one that is treated as a reading.
-
It is an estimate, not a result. Prepyo has no access to Pearson or IDP marking, and no practice platform can issue an official score.
-
Human examiners vary. Two IELTS examiners can place the same essay half a band apart, so treat any single score — ours or a tutor’s — as a reading, not a verdict.
-
Unusual answers are where estimates are weakest. A response that is strong but far outside the expected shape of the task is the case any automated marker handles least well.
-
The number is the least useful part of the report. What moves your band is the criterion breakdown telling you where the marks went, which is why it sits above the score rather than below it.
See it on your own writing
The free diagnostic takes about twenty minutes and returns a breakdown across all four skills, marked the way this page describes.