IELTS Mentor AI adds Reading and Listening — what changed

IELTS Mentor AI has covered Speaking and Writing since launch. As of today, it also runs a full Academic Reading paper and a full Listening test, each marked the moment you submit. This isn’t a features list — it’s a look at what actually changed, what it looks like in the app, and why two of the choices behind it (evidence highlighting, and telling you which mode you just sat) matter more than they might sound.

What a full Reading paper looks like now

A Reading session is a complete Academic paper: three passages, rising in difficulty, 40 questions numbered continuously from 1 to 40 rather than restarting at each passage — the detail that instantly signals a real IELTS layout rather than a generic quiz. You can sit all three passages under the real 60-minute limit, choose exam timing or work untimed, or pick just the passage or passages you want to practise; with exam timing on, the test submits itself automatically when time runs out, the way the real one does.

Marking is objective and instant — no AI in the loop, because Reading answers are either right or wrong against a key. Results come back with a band, the time you used, a breakdown by passage, a breakdown by question type, and a note on which question type cost you the most marks, so you know where to start your review.

Evidence highlighting: why it matters for self-study

The single most common complaint about self-study Reading is not the question — it’s the answer key. You get a question wrong, look up the correct letter, and are no closer to understanding why the passage supports it. A book can’t fix that. A PDF answer key can’t either.

Reading review here does: each wrong answer links to “Show the evidence in paragraph X,” and tapping it opens the passage, scrolls to that paragraph, and highlights the exact sentence the answer turns on. On a straightforward multiple-choice question that’s a nice-to-have. On True/False/Not Given and Yes/No/Not Given — the two types candidates most often get wrong, because they demand telling “the passage contradicts this” apart from “the passage never mentions this” — it’s the difference between guessing better next time and actually learning the skill.

What a full Listening test looks like now

A Listening session runs the real four-part structure: an everyday conversation, then an everyday monologue, then an academic discussion, then an academic lecture, ten questions each, forty in total, each part harder than the one before. Speakers carry different accents — the app itself states British, Australian and American — because meeting an unfamiliar accent for the first time on test day is its own kind of surprise. The scripts include the traps that actually cost marks in the real test: a speaker correcting themselves mid-sentence, a name spelled out letter by letter, a question that says “ONE WORD ONLY” and means it.

Exam mode or Practice mode — and the app says which

Two modes exist, and the honest thing is that they don’t pretend to produce the same score. Exam mode plays the recording once, with no pausing or rewinding. You get thirty seconds of reading time before each part, questions for the next part appear automatically as the recording moves on, and a two-minute check window at the end before it submits itself. Practice mode lets you pause and replay as you go, which is a genuinely useful way to work through the material — but a score earned that way isn’t a score earned under exam conditions, and the app doesn’t blur the two. The results screen states which mode you took, right at the top, so a Practice-mode result never quietly gets read as an exam one.

“Hear it again”: the same idea, for your ears

Evidence highlighting answers “why was this the answer” for Reading. Listening review answers the equivalent question for your ears: for every question, Hear it again replays the exact line the answer came from, and Transcript opens the script with that line highlighted, speakers labelled by accent. Finding the fifteen seconds of a thirty-minute recording that a wrong answer came from used to be tedious enough that most self-study candidates simply skipped it. Here it’s one tap.

One Skill Map, fed by all four skills now

The Skill Map — a set of language nodes shown as percentages, with the weakest one flagged as the skill to fix first — used to run on what Speaking and Writing sessions revealed about your grammar, vocabulary and structure. It’s one map fed by every test you take, and that now includes Reading and Listening: a session in either now moves the same nodes a Speaking or Writing session does, instead of sitting to one side as an unconnected score.

What hasn’t changed: Speaking and Writing

Speaking still runs a full 3-part test — about fifteen minutes in the app — and gives you a band with feedback, including the four criterion bands. Topic Speaking is unchanged too: Part 1, Part 2 and Part 3 tabs, pick a category then a topic, and an AI examiner that asks follow-up questions rather than reading from a fixed list; prepare for sixty seconds, then speak for up to two minutes.

Writing still marks Task 1 (Academic and General Training) and Task 2 essays with live AI marking: an overall band, the four criterion bands, an Examiner’s Summary, what worked and how to improve for each criterion, and a Data Accuracy card that separates a misreading of the chart from a language mistake. The editor is still timed — 20:00 for Task 1, 40:00 for Task 2 — with a live word count against the 150/250-word minimums, and you can still attach photos of a handwritten answer instead of typing it.

Academic and General Training, precisely

Worth being precise about, since the two tracks diverge: Reading in the app is Academic only — three passages of the kind an Academic candidate sits, not the General Training format. Writing Task 1 covers both Academic and General Training prompts. If you’re not sure which of the two you’re taking, that distinction is worth settling before you sit a Reading paper here.

Where this fits

None of this replaces knowing your target. If you’re weighing how a Reading or Listening band actually rolls into your overall score, the band score calculator converts raw scores for both and shows the arithmetic. If Speaking is still the skill you’re focused on, how IELTS Speaking is scored covers the four criteria the same way this page covers Reading and Listening, and the roundup of IELTS Speaking apps is still accurate for that comparison. For the two skills that are new here, the Reading practice apps worth knowing about and the Listening practice apps worth knowing about put this update next to what else exists, honestly, including where the others do things this app doesn’t.