SCORING GUIDE
How the 0–75 score maps to a CEFR band
Every Multilevel speaking attempt returns a raw score out of 75. That score maps directly to a CEFR band, using the same thresholds as the official DTM/UZBMB rubric.
The band thresholds
| Score | Band |
|---|---|
| 65–75 | C1 |
| 51–64 | B2 |
| 38–50 | B1 |
| 20–37 | A2 |
| 0–19 | A1 |
Want the full picture of what separates each band? See the CEFR bands page.
What each part is assessed on
LimeTalk grades each part holistically — one band per part, not a checklist of sub-scores — using the descriptors examiners actually apply:
- Grammatical range & accuracy
- Vocabulary
- Fluency
- Coherence & cohesion
Pronunciation isn't scored as a separate criterion — it isn't judged from a transcript, so it doesn't factor into the holistic band the same way grammar, vocabulary, fluency, and coherence do.
How much each part is worth
Parts aren't weighted equally. Part 1.1, Part 1.2, and Part 2 are each scored on a 0–5 scale; Part 3 is scored on a 0–6 scale. Out of the raw 21-point total, that gives Part 3 the single biggest share of any one part:
| Part | Raw scale |
|---|---|
| Part 1.1 | 0–5 |
| Part 1.2 | 0–5 |
| Part 2 | 0–5 |
| Part 3 | 0–6 |
The raw 0–21 total is then converted to the headline 0–75 score using the official conversion table — it isn't a straight multiply-by-3.57, so a single strong or weak part moves the final score by more or less depending on where the total lands.
How each part targets a different level
The four parts step up in difficulty on purpose, so your weakest part usually points at your real ceiling:
Is the score official?
No — LimeTalk is an independent practice tool. Scores and bands are AI-generated estimates meant to mirror the real rubric closely enough to be useful, not a substitute for your actual DTM result.
See how you score against the real rubric.
Take a full test