IB Math AI IA Grader

Grade your Math AI IA before you submit. Upload your draft and get a mark for each criterion, comments tied to your sentences, and a list of what to fix first.

5 free gradings a month. No card. Private to you.

Math AI IA mark and criterion feedback in Marksy
of marks within one grade of the IB's final mark
92%
marking score. ChatGPT 65, Claude 61, Gemini 57
74/100
real IAs and TOK essays tested against the mark the IB gave
1,317
Marks
20, from five criteria
Length
About 12 to 20 pages
Share of your grade
20%, at SL and HL
SL and HL
Only criterion E is different

See the mark, and the sentences behind it

Check the quality of your method, analysis, evaluation, communication and subject-specific reasoning. Find the weakest parts of the draft and see what to change first.

Math AI IA: comments on your writing
Comments on your writingEach comment points at a passage in your draft, so you can see why Marksy said it.
Math AI IA: what to fix first
What to fix firstA short list of edits in order, instead of changing random parts of the draft.

What Marksy checks for each Math AI IA criterion

You get a mark for each IB assessment criterion below. The points under each one are what Marksy looks for in your draft. See the mark bands

Criterion A: Presentation

4 marks
  • The exploration has a clear aim and a logical sequence.
  • Graphs, tables, diagrams, and calculations appear where the reader needs them.
  • Repetition and irrelevant material do not obscure the mathematical story.

Criterion B: Mathematical Communication

4 marks
  • Variables, notation, units, and important terms are defined and used consistently.
  • Equations and representations are formatted and explained clearly.
  • Calculator or software notation is not used in place of proper mathematical notation.

Criterion C: Personal Engagement

3 marks
  • Choices of question, method, examples, or technology have a clear rationale.
  • You test ideas or make decisions instead of following a fixed template.
  • Engagement is shown through the work, not claimed in an isolated statement.

Criterion D: Reflection

3 marks
  • Reflection appears throughout the exploration rather than only in the conclusion.
  • Results are checked for reasonableness, limitations, and alternative interpretations.
  • You explain what the findings mean for the original aim.

Criterion E: Use of Mathematics - SL

6 marks
  • The mathematics chosen is relevant to the aim and at the level of your course.
  • Work is correct, with technology outputs interpreted rather than pasted.
  • Explanations show why the methods work and what the results mean.

Criterion A: Presentation

4 marks
  • The exploration has a clear aim and a logical sequence.
  • Graphs, tables, diagrams, and calculations appear where the reader needs them.
  • Repetition and irrelevant material do not obscure the mathematical story.

Criterion B: Mathematical Communication

4 marks
  • Variables, notation, units, and important terms are defined and used consistently.
  • Equations and representations are formatted and explained clearly.
  • Calculator or software notation is not used in place of proper mathematical notation.

Criterion C: Personal Engagement

3 marks
  • Choices of question, method, examples, or technology have a clear rationale.
  • You test ideas or make decisions instead of following a fixed template.
  • Engagement is shown through the work, not claimed in an isolated statement.

Criterion D: Reflection

3 marks
  • Reflection appears throughout the exploration rather than only in the conclusion.
  • Results are checked for reasonableness, limitations, and alternative interpretations.
  • You explain what the findings mean for the original aim.

Criterion E: Use of Mathematics - HL

6 marks
  • The mathematics chosen is relevant to the aim and at the level of your course.
  • Work is correct, with technology outputs interpreted rather than pasted.
  • Explanations show why the methods work and what the results mean.

Where Math AI IAs usually lose marks

Marksy flags these in your draft and links each one to the sentences it is about.

  • Raw data in the middle of the workPages of raw data break up your analysis. Show a short sample where you use it, and put the full data set in an appendix.Criterion A
  • Spreadsheet output with no notation"R² = 0.87" copied from a spreadsheet isn't enough. Define each variable with its units, write the model as an equation, and round the same way throughout.Criterion B
  • Data with no decisions in itUsing a downloaded data set is fine. Engagement shows in what you decide: which data to keep, how you handle outliers, and which model to try next and why.Criterion C
  • One model, never testedWhere your data allows, fit more than one model and compare them with real values. Say where the best one breaks down, such as outside your data range.Criterion D
  • Misreading the statisticsAn r of 0.8 doesn't mean 80% accurate, and a strong correlation doesn't prove a cause. Say what each value tells you about your question, and set your significance level before you test.Criterion E
  • A test without its conditionsA chi-squared test is unreliable when expected frequencies are small. Check the conditions for each test you use, and say what you changed when one wasn't met.Criterion E

Before you upload

  • Say where your data came from, and how you collected or cleaned it.
  • Write equations, r values and p-values in the text, not only in screenshots.
  • Label every graph and table, and refer to each one in your writing.
  • Choose SL or HL when you upload, because criterion E changes.

Your draft stays at the centre

Marksy reads the work you upload, points each comment back to your writing, and puts the most important changes first.

Does the mathematics fit the context, and does the context actually matter to the mathematics?
Read Mathematics: Applications and Interpretation IA Planning Guide
  1. 1

    Upload your draft

    A PDF of your Math AI IA, finished or not. Every comment stays connected to your own writing.

  2. 2

    Read the comments

    Highlights show which passage each comment is about.

  3. 3

    Fix the important parts first

    Work down a short list instead of making scattered edits.

  4. 4

    Keep the final say

    Accept, adapt or ignore each suggestion. You stay the author.

What students wrote afterwards

From Trustpilot, as written.

marksyib has saved my ib diploma. thanks to it, my IA's were perfect and got the points i needed. really recommended.
maya balog, 13 February 2026
Marksy ib allowed me to be able to finish my IAs much before allowing me to be able to relax better. 100% recommended
Jordi Pasalaigua Peiron, 25 March 2026
Read the reviews on Trustpilot

Your work stays your work

Marksy reads the draft you upload and comments on it. It doesn't rewrite your assessment, and your files are never used to train AI.

Read the safety answers
  • Private to youOnly you can see your work unless you choose to share it.
  • Never used for trainingYour coursework is not added to any AI training data.
  • Feedback, not ghostwritingUse the report to improve your own analysis and decisions.
  • Delete it whenever you wantYou control the files stored in your account.

Questions before you upload

Still planning? The coursework guide covers the criteria and what to ask yourself while you write. Open the guide

Yes. Say where it came from and why you trust it. Show the choices you made with it, such as removing outliers, because those choices are where your own thinking shows.

It can be, if the statistics come from your course, are correct and you explain why each method fits. At HL, criterion E looks for sophistication and rigour, so tests you also meet at SL rarely carry it alone.

Criteria A to D are the same. Criterion E judges your maths against the level of the Math AI course you take, SL or HL. Math AA has its own grader page.

Along with the IB criteria, Marksy's maths marking uses notes on how real IB-marked explorations scored at each mark band. That includes common slips with regression, correlation and hypothesis tests.

We tested Marksy on 1,317 real IAs and TOK essays with the mark the IB gave them. 92% of Marksy's marks landed within one grade. On a set with as many weak as strong essays, it scored 74/100 on our marking score, against 65 for ChatGPT, 61 for Claude and 57 for Gemini. Every number, and how we tested, is on the grading accuracy page.

Yes. Marksy checks your Mathematics: Applications and Interpretation IA against the IB assessment criteria for it and returns marks, comments and fixes instead of generic essay advice.

Yes. Early drafts are useful because the report can show a weak criterion before you spend hours polishing the wrong section. You can revise and grade another version later.

A free account includes 5 gradings a month. You don't need a card.

No. Marksy reads your existing work, explains each comment, and suggests what to revise. The writing and the final decisions stay yours.

Your upload is private to your account, is not used to train AI models, and can be deleted whenever you want.

Stop guessing what to fix in your Math AI IA

Upload the version you have now and find the parts holding it back.