What Bandzen is

Bandzen is an IELTS preparation product. You practise on real question types, your work is marked, and what the marking finds decides what you are asked to do next.

What is built

Reading and Writing are the two modules with engines behind them.

Reading is marked by a pure function — no model is involved, the answer key is fixed, and the same attempt would score the same every time. Writing is marked by a language model against the IELTS band descriptors, and the report names four criteria, quotes your own sentences back to you, and says what to fix.

Everything else follows from those two: your estimated band, your study plan, the skill matrix on Progress, and what Coach knows about you.

What is not built

Not built

Listening needs audio nobody has recorded yet. Speaking needs recording and transcription that has not been built. Both appear in the navigation as locked states saying exactly that, and neither can be unlocked by paying.

There is also no full four-skill mock test, for the same reason — two of the four skills do not exist to mock. The diagnostic is the closest thing: one reading passage and one Task 2 essay, sat under timing, in one sitting.

Reading practice is currently marked on the Academic conversion scale for both Academic and General Training candidates. General Training converts differently and more generously; until that ships, a General Training candidate should read their reading band as conservative.

Estimates, not scores

Every band Bandzen shows you is an estimate produced by this product. It is not an IELTS score, it does not come from IELTS, and no one here can tell you what you will get on the day.

Bandzen is an independent IELTS preparation platform and is not affiliated with or endorsed by IELTS, the British Council, IDP, or Cambridge University Press & Assessment. Band estimates provided by Bandzen are for practice purposes and are not official IELTS scores.

The product holds itself to that everywhere. Progress shows no projected band and no "you could reach 7.0" line, because there is no basis on which to promise a gain. Activity counts minutes spent inside timed attempts only, because nobody measures how long you spend reading a lesson, and an estimate there would be a guess dressed as a figure.

Where to go next