Assessment scope
The assessment covers Listening, General Reading, Letter Writing, Essay Writing, Speaking. It reports only within this stated scope and does not infer unmeasured abilities.
Adaptive selection and coverage
The shared engine chooses the next eligible item from the track question bank while balancing coverage, current evidence, and the need to verify an uncertain boundary. A fixed seed makes the same technical scenario reproducible in automated tests.
Personalization questions describe goals, time, or preferences. They shape the explanation and plan but do not enter the knowledge score.
Quick and Full
- Quick uses up to 20 questions and provides a directional reading.
- Full uses up to 40 questions and gathers broader evidence.
- Both modes use the same track contract and scoring model.
- The report states the mode so a shorter result is not presented as more certain than its evidence.
How the result is formed
The engine summarizes the overall outcome, dimension outcomes, strongest dimension, primary gap, and any boundary or low-confidence condition. Internal estimates and numerical confidence remain hidden from the visible report and share payload.
Storage, sharing, and limits
- Produces no band score, predicts none, and offers no equivalent range.
- Determines no immigration or visa eligibility and no authority's requirement.
- Plays no audio, so it does not measure listening performance.
- Assesses no free writing and records no speech.
- Does not measure speed under time pressure or reproduce the official section order.
Official sources
Links should be rechecked before a high-stakes decision.
Choose your assessment depth
Use Quick for a directional reading or Full when you want broader evidence and a richer report.