Skip to content

Second-pass marking — how it works

MarkScheme is built for scheme-aligned feedback after you self-mark: photograph handwritten Cambridge or IB answers, get B1/M1/A1 or markband breakdowns in ~30 seconds, then redo one skill before the next paper.

How it works

From photo to marked script, in four honest steps.

No magic claimed. Here's exactly what happens to your work — and where the limits are.

MarkScheme marking result showing 3 out of 3 marks, official 9709 mark scheme banner, and the binomial expansion question
Step 1

Your score, right there

Marks earned, the paper, and the question — checked against the official Cambridge scheme for that session.

Mark-by-mark breakdown with B1, M1, and A1 earned explanations for a Mathematics answer
Step 2

Every mark, accounted for

You'll see exactly where you nailed it and where you slipped — B1, M1, A1 explained like an examiner would write them.

Examiner's Ink on handwritten binomial expansion work with B1, M1, and A1 stamps on the student's script
Step 3

Ink on your paper

Stamps and notes sit on your actual handwriting — not a vague paragraph at the bottom of the page.

Worked solution prompt and buttons to mark another attempt or a new question
Step 4

Then learn and retry

Open a worked solution when you want the method, then mark another go at the same question or move on.

Trained on real Cambridge mark schemes. Returns red-pen annotations and structured feedback in under a minute. Five questions free, no card.

Adaptive marking

MCQ keys, B1/M1/A1 step marks, and essay band descriptors — the engine detects the question type and marks it the way Cambridge does, not with a generic rubric.

Examiner's Ink

Red-pen-style notes anchored to your actual handwriting. See which line earned B1, where M1 was lost, or why your essay sits in a particular band — not a vague paragraph at the bottom.

Mastery tracking

Each attempt feeds your syllabus coverage map. Spot blindspots, track progress across spec points, and know what to revise next — built from your real marking history.

Honest about the AI

What it does well — and where it's limited

What it does well

Applies the official scheme criteria, mark by mark, with citations
Reads most handwriting, including multi-page working
Spots recurring error patterns across your attempts

Where it's limited

Genuinely illegible lines are flagged, not guessed
Essay band judgements are approximate — a human examiner may differ
Grade estimates are boundary-pattern approximations, not predictions
Try it on one question

FREE TIER · ABOUT A MINUTE · NOT ENDORSED BY CAMBRIDGE INTERNATIONAL