Why does sayit's IELTS band estimate stop at 8.5, and what does it actually miss?

sayit's IELTS Speaking band estimate is capped at 8.5, and that ceiling is a deliberate, honest limit rather than a technical shortfall to apologize for: it comes from measuring only two of the four official IELTS Speaking criteria — Fluency and Coherence, and Pronunciation — while Lexical Resource and Grammatical Range and Accuracy, the two that increasingly separate the top of the scale, aren't assessed at all.

30-second version: A real IELTS Speaking band is the average of four criteria scored by a human examiner. sayit's estimate only measures the two that can genuinely be heard from audio alone — fluency and pronunciation — and blends them into a number labeled clearly as an estimate, never as an official score. Because the two unmeasured criteria matter most at the very top of the scale, the estimate is capped below a perfect score rather than pretending it can see what it can't.

Which two criteria does the estimate actually measure?

Fluency and Coherence — your pacing, pausing, hesitation, and how smoothly your speech flows — and Pronunciation, the accuracy of your individual sounds, your word stress, and your intonation. Both are audible properties of a recording, which is exactly why an audio-only system can assess them with real confidence. The specifics of how each is measured are covered in fluency and pacing and pronunciation scoring.

Which two criteria does it not measure, and why not?

Lexical Resource — the range and appropriateness of your vocabulary — and Grammatical Range and Accuracy — the variety and correctness of your sentence structures. Neither of these is fundamentally about how something sounds; they're about what was said and whether it was said correctly in a linguistic sense, which requires understanding meaning and grammar, not just acoustic signal. sayit doesn't score grammar or vocabulary range anywhere in the product, on IELTS prompts or otherwise — this isn't a gap unique to the exam feature, it's a boundary the whole scoring engine respects.

Why 8.5 specifically, and not some other ceiling?

Because the two unmeasured criteria matter most at the very top of the real scale — the difference between a 7 and a 9 on the actual exam is driven substantially by vocabulary sophistication and grammatical range, exactly the dimensions this estimate can't see. Reporting up to 8.5 acknowledges that a genuinely strong speaker with excellent fluency and pronunciation deserves a high estimate, while refusing to claim the very top of the scale, where the unmeasured criteria would be doing real, decisive work.

Does the estimate call itself "official" anywhere?

No — every estimate carries an explicit label that it is not an official IELTS score, alongside a short caveat explaining what it actually measured. This matters most for anyone using it to prepare seriously for the real exam: a strong estimate here is a genuine signal about your fluency and pronunciation, not a guarantee of what a human examiner would award once vocabulary and grammar enter the picture.

What the estimate covers, at a glance

So what is the estimate actually useful for?

Tracking real, measurable progress on the half of speaking performance it can see. If your fluency and pronunciation are the bottleneck — hesitation, dropped consonants, flat intonation — the estimate will move as those genuinely improve, and the examiner-style notes attached to each attempt tell you specifically what to work on. It's a real diagnostic for a real part of the exam; it was never meant to replace the other half.

Try it

Answer a prompt at IELTS speaking in sayit and read the estimate alongside its caveat — the number is honest about what it saw, and about what it didn't.